Chapter III: The Somme Battles

I.

As the companies of the Heavy Section arrived in France, they were assembled with their tanks at Yvrench, near Abbeville. This was before the days of the Channel Ferry. The heavy machines had to be shipped like any other form of ordnance; and they were but a part of the load which required to be carried overseas. Several complete workshops, dozens of sponson trolleys, light cars, lorries, and motor-bicycles, together with tons upon tons of spare parts, furnished the first instalments of that vast collection of material which the Tank Corps came to accumulate in France.

A training centre was established at Yvrench, at first under Lieut.-Colonel Brough, but early in September Colonel Bradley came out to take over the command. It was known that the unit would soon be called upon, and there still was an immense amount of work to be done. Drivers and gunners had to receive further training. The tanks required to be overhauled and tuned up. Workshops and depots must be established. Above all, a host of novel problems concerning administration, supply, tactics--in short, the whole method and manner of tank-work in the field--fell to be solved without delay, and, what was more, without assistance. The tank companies in France were nobody’s children. Their nominal parent, the M.G.C. (itself a mushroom growth of doubtful antecedents and future), had in fact no actual connection with them whatever; and it would seem that a new technical unit, in so experimental a stage, should more properly have been made a branch of an established technical corps, such as the Royal Engineers. As it was, the companies were self-contained and unattached. The infantry, whom they had come to help, knew nothing of them except by rumour. The Generals and staff officers who came to see them perform regarded them as a curious and rather dangerous innovation, unlikely to be of much practical value. In so far as this isolation was intentional, it is very doubtful whether the measure of secrecy obtained by it outweighed its disadvantages. The first battle was near at hand; and a few minutes after zero the secret would be exploded for ever. In the meantime, a fuller comprehension by other arms, a more liberal interchange of views, in especial some tuition of the infantry who were to co-operate in action, might have saved much disappointment and many lives. Nothing of the kind was attempted; and the companies had to think out and master other people’s problems in addition to their own, in the field and with small aid from outside. Some contingencies inevitably were overlooked. Not only was experience of this new warfare yet to seek, but many of the officers and a larger proportion of the men actually were in France for the first time, and knew nothing of normal battle conditions. It is surprising how much was foreseen and provided for in spite of these disadvantages. The matter of supply, for example, seems to have been dealt with adequately from the beginning, for the tanks were seldom hampered for want of petrol or ammunition, wherever they might be. As regards general tactics, the new unit was fortunate in possessing a basis to work on in a valuable memorandum issued by Colonel Swinton early in the year. In this he outlined the probable scope of tank warfare, dealt with infantry co-operation and artillery support, and laid down a definite system of attack which in essentials is in use to-day.[6] On the other hand, less thought was given to a Department that proved to be of vital importance. This was Intelligence, or (as the branch of that service in the Tank Corps is more accurately styled) Reconnaissance. Circumstances peculiar to tank operations call for the training of special officers in this duty alone, as well as an unusually high standard of proficiency in all tank commanders; but these needs were not recognised fully at the outset, and one Intelligence officer, Captain Hotblack, was responsible for the welfare of all four companies. A little later, and there were seven Reconnaissance Officers to each battalion; and the corps soon won so high a reputation for thoroughness in this work that every formation in the Army, from Headquarters downward, came to it for assistance.

II.

In the beginning of September the situation along the Somme Front was as follows:--

After two months’ very sanguinary fighting two stages of the battle had been completed, more or less satisfactorily. In the British sector, which alone concerns us, the enemy’s first and second positions had been captured from a point just west of Combles (where we touched the French 6th Army) to Pozières on the Albert-Bapaume road. Farther northward, from Pozières to Gommecourt, we had failed badly, and were still in our old front line. The casualties had been extremely heavy. Before Beaumont Hamel and Thiepval, whole brigades were destroyed on the first day without gaining a yard of ground; and along this moiety of our front the battle came virtually to a standstill for months. In the southern portion our advance, although slow and costly, had progressed fairly methodically, side by side with that of the French. The result was a flattish salient biting into the enemy’s old position from the Albert-Bapaume road, southward across the Somme, to the road from Amiens to Brie. Corresponding to the curve of the salient was the German third line of defence, completed since the battle began. On the British front it ran along the reverse slope of the Pozières Morval ridge (most of which was in our possession), some five miles west of Bapaume. It was now intended to resume our advance in an easterly and north-easterly direction toward that town. The tunnelled positions of Thiepval and Beaumont Hamel were to be left, for the present, still unassailed, except for a few minor enterprises required to improve the existing line.

The main attack was fixed for the 15th of September. Both the British armies now engaged in the general offensive were taking part: the 4th, under General Rawlinson, whose front extended from the point of junction with the French to west of Martinpuich, and a part of Gough’s 5th or Reserve Army, which continued the line northward. From right to left the 4th Army was to carry the villages of Morval, Les Bœufs, Gueudecourt, and Flers; while Gough, after a preliminary operation south-west of Thiepval on the 14th, had in hand the capture of Martinpuich and Courcelette. The usual orders were issued for a vigorous exploitation of any opening--a poor hope on the Somme--and once more the Cavalry Corps, with all its impedimenta, clattered noisily up toward the battlefield to be at hand if needed.

The tanks were to make their entry in this attack. C and D Companies, commanded respectively by Major Holford Walker and Major Summers, D.S.C., were allotted to the two armies, and on the 10th moved up to the railway centre called the Loop, near Bray. From there the tanks trekked to their lying-up positions for the battle--C to the Briqueterie near Trônes Wood, and D to the Green Dump behind Delville Wood. The companies’ headquarters followed on the 14th. For the actual attack the tanks were to be distributed as follows:--

Three sections of each company (17 tanks apiece) worked on the right and right-centre with the 14th and 15th Corps. One section of 8 tanks of D Company was attached to the 3rd Corps in the centre, while the remaining section of C (7 tanks) came under the orders of the Reserve Army on the left. This made a total of 49. It was intended for the machines to be used in sub-sections of two or three against specified strong-points. There was some discussion as to whether they should precede or follow the infantry at zero. If they led, it was feared that the noise of their approach to the front line would bring down the enemy’s barrage upon our assembly trenches; and eventually it was decided that their arrival should be so timed that they could reach the first objectives--roughly the line of villages--five minutes before the infantry. For this purpose lanes were to be left in our own barrage.

I shall not attempt a detailed account of this or any other early tank action. In the first place, it would be difficult to do so, for the records of those days are meagre and inaccurate. Again, the _rôle_ of the tanks on the Somme was so largely experimental and subsidiary that infantry operation orders, even after the secret was blown, were framed to be carried out irrespective of the action of what was then regarded in most quarters as an unreliable innovation of doubtful future. As regards the tactical methods employed, an organisation in its infancy, with everything to learn, had to take what chances it was offered, and had little say in the direction of events. No attempt was or could be made at this stage, or for long after, to introduce those revolutionary tactics (already discussed and advocated by Colonel Swinton and others) in which the tank was to be the vital instrument, and wherein lies the chief interest of its history. Finally, since all tank battles as such bear very similar features, I prefer to concentrate on later ones of which I have personal knowledge.

The attack on the 15th September was highly successful on the left and centre, but failed on the right, where the strong-point known as the Quadrilateral held up our advance against Morval and Les Bœufs. On the whole, it may be said that the work of the tanks fully justified expectations. No startling triumph could be recorded, but under the circumstances none was anticipated. Enough was done to demonstrate the inherent soundness of the design, the correctness of the tactical axioms laid down, and the enormous potential value of the machine when faults had been rectified and the personnel more highly trained. The test to which the Heavy Section had been put was a severe one. The ground everywhere was so pulverised by the ferocious shelling of the past two months, that out of 49 tanks which left the Briqueterie and Green Dump that night, 17 failed to reach their starting-points. These broke down or were ditched on the way, the absence of any unditching gear, such as was supplied later, proving a serious handicap. Half an hour before zero, which fell at dawn, the other 32 began to crawl forward to the front line. The three sections of C Company, on the right, had been particularly unfortunate during the approach-march, and the few tanks remaining were able to effect little or nothing against the Quadrilateral and Les Bœufs. In most cases they were ditched or knocked out by shell-fire soon after they crossed the German front line. D Company’s three sections, working with the 15th Corps farther north, were far more successful. Every one who read the newspapers at the time knows how one tank preceded the New Zealanders up the main street of Flers, driving the terrified Germans down into the cellars, where the infantry rounded them up. The tank then moved on to a trench strongly held by the enemy, enfiladed it, moved along behind it, and took 300 prisoners. Farther north, the two sections assisting the 3rd Corps and the Reserve Army reached Gueudecourt and the famous sugar factory at Courcelette. In almost every case where a tank succeeded in arriving at its objective, the mere sight of it was enough. Save for some picked machine-gunners and a few other desperate men, the enemy fled or surrendered on the spot.

The moral effect, all proverbs about familiarity to the contrary, is still and will always remain the chief asset of the tank as a weapon. No body of infantry, unless very closely supported by guns, will stand for long in the open before a properly concerted attack led by tanks. They cannot: accidents apart, they are entirely helpless, and they know it. The fire from the machine may be erratic, as usually it is, for the gunners can see very little, and their platform is continually in motion; but this does not abate in any degree the moral effect of its onset. In the last five months of the war, when the proper use of tanks had come at last to be understood, it was only by the bold use and sacrifice of field artillery, pushed far forward, that the German armies were able to protract for as long as they did the final struggle. And this moral influence works both ways: it helps the attacker as much as it demoralises the attacked. The tank is terrifying, or inspiring and rather ludicrous, according to whether it is against you or for you. It is not only a known antidote to the dreadful trinity of trench, wire, and machine-gun. Its very appearance is a comfort. Its shape, unlike that of any other example of man’s handiwork, its deliberate but at times curiously active and disconcerting movements, its obvious weight and power, even the two ports, like lidded eyes, which ornament the front of the cab, combine irresistibly (as all journalists will testify) to suggest some huge prehistoric animal--some giant sloth or toad. Indeed, to visit a training-ground in the evening when some dozens of these creatures are rolling home to their stalls or hangars is, for all the world, like being at a Zoo during the Pleistocene Age. The striking pattern of colours with which the machines were painted in the early days only aided the illusion. The British soldier, whose great merit it is to see the comic side of everything, could not fail to be exhilarated by these gorgeous monsters. It was long before he came to rely on them as he could rely, let us say, on an old-established and perfected arm like the artillery; but this is not remarkable. Rather is it something of a miracle, and an abiding testimony to the efforts of the new corps, that a similar confidence between infantry and tanks actually was created during the war’s lifetime. And from the very first their mere appearance in the field was a great moral factor in our favour.

In this, their first battle, their casualties in personnel were insignificant. One tank, after attacking a field battery in Gueudecourt and destroying one gun, was itself hit and burnt out, only two of the crew escaping. Most of the others got off lightly. Altogether ten were disabled by shell-fire, and seven more were hit, but returned to the rallying-point.

Reserving remarks and criticism for the end of the chapter, we will run through the remaining actions fought by the Heavy Section that autumn and winter. A comparative lull of rather more than a week (marked by a successful minor operation which gave us the Quadrilateral) followed on the battle of the 15th; but during this period the tank companies were feverishly busy. At first no fewer than 50 tanks were asked for to assist in the next attack. So large a demand at that stage meant an immense amount of work; and in the end nothing like this number was required, the co-operation of the Heavy Section being limited to the northern assault on Gueudecourt, Thiepval, and the height overlooking the Ancre from the south. Work on the tanks was hampered by the multitude of visitors who came to see these novelties now that the secret was disclosed. There was much uncertainty as to the method of attack, the supply of maps was insufficient, and operation orders, when they did arrive, were meagre and ambiguous.

The ground along the Thiepval ridge was in an appalling condition. It had been shelled more or less continuously for upwards of three months by guns of every calibre. On top of this came two or three days of heavy rain just before the attack. “The plan,” to quote Colonel Buchan, “was for an attack by the 4th Army on Monday, the 25th, with--on its left wing--small objectives; but, on the right and centre, aiming at completing the captures which had been the ultimate objectives of the advance of the 15th. The following day the right wing of the 5th Army would come into action, and it was hoped that from Thiepval to Combles the enemy would be driven back to his fourth line of defence, and our own front pushed up well within assaulting distance.” The enemy’s fourth line of defence ran just in front of Bapaume. Five tanks only were detailed eventually to work with the 4th Army on the 25th, and they were able to effect little or nothing. Unable to leave their starting-point until zero, for the attack was made at midday, they seem to have become ditched very early. They were hardly needed, except at one point. The assault was entirely successful everywhere but at Gueudecourt, where uncut wire held up the 21st Division. It had been a bitterly disappointing day for the tanks, but the fault was hardly theirs. It was an error in the first place to employ them at all in an operation starting in broad daylight, for it was impossible to conceal them near enough to the front line to enable them to go over with the infantry. Corps and divisional commanders had not yet learnt the elements of tank warfare. Secondly, the ground was so riven and pulverised, so absolutely rotten with shelling and rain, that without any unditching gear or spuds (iron shoes afterwards provided to grip sodden or crumbling soil) the odds were all against their getting far.

As it happened, an opportunity was soon offered of redeeming these mishaps, and was seized in a very brilliant manner. The 21st Division, we have seen, was brought to a standstill in front of Gueudecourt. Here was a formidable obstacle known as Gird Trench, and as a preliminary to a further assault on the village it was essential that this should be cleared. Orders were issued that night for an attack at dawn, and the help of a tank was asked for. With immense toil one was brought up in the dark across the frightful waste of half-obliterated trenches and shell-holes and water. For the ensuing action I will quote the report of the 15th Corps.

“On the 25th September, the 64th Brigade, 21st Division’s
attack on Gird Trench was hung up and unable to make any
progress. A footing had been obtained in Gird Trench at N. 32 d
91, and our troops held the trench from N. 26 c 45 northwards.
Between these two points there remained approximately 1500
yards of trench very strongly held by the Germans and well
wired, the wire not having been cut. Arrangements were made
for a tank (female) to move up for an attack next morning. The
tank arrived at 6.30, followed by bombers. It started moving
south-eastward along the Gird Trench, firing its machine-guns.
As the trench gradually fell into our hands, strong-points were
made in it by two companies of infantry which were following
in rear for that purpose. No difficulty was experienced. The
enemy surrendered freely as the tank moved down the trench.
They were unable to escape, owing to our holding the trench at
the southern end at N. 32 d 91. By 8.30 A.M. the whole length
of the trench had been cleared, and the 15th D.L.I. moved over
the open and took over the captured trench. The infantry then
advanced to their final objective (Gueudecourt), where the tank
rendered very valuable assistance. The tank finally ran short
of petrol south-east of Gueudecourt. In the capture of Gird
Trench 8 officers and 362 other ranks were made prisoners, and
a great many Germans were killed. Our casualties only amounted
to 5. Nearly 1500 yards of trench were captured in less than
an hour. What would have proved a very difficult operation,
involving probably heavy losses, was taken with the greatest
ease entirely owing to the assistance rendered by the tank.”

It should be added that an aeroplane also lent a hand, flying low over the trench and machine-gunning the defenders.

This admirable little demonstration of what a tank could do came at a most opportune moment. The months of slaughter on the Somme were in fact having their inevitable effect on the armies in France. Already the best of the volunteers who enlisted in the first year of the war were gone, and the new recruits were of a very different type. Again and again battalions, brigades, whole divisions even, came out of the line decimated and shaken from what was trumpeted at home as a cheap and glorious success. A mile or two of trenches, a wood or a village, had indeed been won, and with much glory. But at what a cost! Whether or no these Pyrrhic victories were worth the sacrifice--and presumably they were--the men on the spot, seeing only their small corner, remembered nothing but the horror of it all. The Somme, together with Gallipoli, and the third battle of Ypres, stand apart in the minds of those who lived through them as long-drawn nightmares; and it was the peculiar abomination of the first-named that it seemed but the foretaste of an endless future. The war might go on for ever in this fashion. There can be no doubt that the Higher Command itself viewed with apprehension a prolonged continuance of these conditions, and was prepared at last to welcome any suggestion or innovation which promised to alleviate them. Hence the demand for the services of the Heavy Section before, in fact, it was prepared to take the field. The tanks came; and the little action just described, together with other similar if less spectacular performances, went far to convert doubters to the possibilities of the attack by armour. Gird Trench, however insignificant in itself, was a singularly perfect example of how the new weapon could help toward economy of life and effort. At a conservative estimate the enemy had lost not only a strong position, but also some 600 officers and men killed, wounded, or taken. Our infantry casualties totalled five, and the tank crew who were directly responsible for the success numbered only eight. Such figures in themselves were convincing.

On the same morning eight tanks went over on the left with the right wing of the 5th Army against Thiepval. The ground was all but impossible, and most of them were ditched on or about the German front line; but two got into Thiepval and gave valuable help to the infantry at a time when the artillery support for once had failed.

III.

The Heavy Section had no more fighting for nearly two months. With October the bad weather set in: the rain fell in torrents: when it did not rain a heavy mist gave the sodden ground no chance of recovery; and the overburdened roads went from bad to worse. While the armies struggled in the mud and water on the slopes of the final ridge before Bapaume, the tank companies were withdrawn to Acheux to refit.

By the end of October it was plain that a condition of stalemate had been reached on the main battlefield. To relieve this it was decided to make a flank attack along and north of the Ancre, where we had been repulsed in July. The capture of the Thiepval height had improved the prospect of a fresh attempt in this quarter. The northern edge of the salient we had created now ran east and west along high ground south of the river; and it was possible for our artillery to enfilade and even take in reverse the immensely strong positions of Beaucourt and Beaumont Hamel.

The attack was fixed for the 13th November. Originally, twenty tanks were to have been used, but the weather was so bad that this number was reduced to five--three to assist at St Pierre Divion, south of the Ancre, while the other two endeavoured to reach Beaumont Hamel. By this time the absolute necessity of good reconnaissance had been recognised; and before the action, tank commanders were taken to observation-points from which they could see something of the ground beyond the front line. Landmarks were indicated to them, aeroplane photographs were studied, and, on zero night, the tank routes were taped as far as possible. But no foresight could avail against the weather. Gales blew, the rain fell, and, on the night of the 12th-13th, came snow. The three southern tanks were bogged in the mud as they moved up. The northern pair found the tapes obliterated by the snow, but Captain Hotblack, the Intelligence Officer, led one forward under heavy fire, and then going back, brought up the second to a point where the infantry was hung up. Both these tanks did good work before getting ditched, and Hotblack received the D.S.O. Meanwhile the attack had succeeded beyond expectation. St Pierre Divion and Beaumont Hamel were taken, although once more an assault on Serre was repulsed. One battalion of the Naval Division, led by Lieut.-Colonel Freyburg, actually pushed right up the river valley to the outskirts of Beaucourt. Here it hung on all day, far ahead of our main advance. The failure to capture a powerful redoubt between Beaucourt and Beaumont Hamel prevented any adequate support reaching the battalion, but Freyburg, who was wounded in all four times, refused to withdraw. A Territorial unit contrived to get in touch with him, and communication of a sort was maintained with the rear along the river bank. But his position was as precarious as it was valuable; and that night three tanks were brought up for another attack on the redoubt. One was hit by a shell and disabled. The others got within close range of the strong-point before getting ditched. They were able to bring their guns to bear, and in a short time the place surrendered, no fewer than 400 prisoners falling to the two tanks alone. The subsequent infantry advance closed up the line and enabled Freyburg to crown his exploit by capturing Beaucourt at the head of his men.

A few days later the battles of the Somme came definitely to an end. If the autumn advance had come up to expectations there was an intention of continuing the offensive throughout the winter, but the powers which modern weapons give to the defence, combined with the October rains, had limited our successes and brought the army to a pitch of exhaustion; and now a complete break-up of the weather put any idea of further operations out of the question. The troops set about making themselves as comfortable as possible in their new positions. Shattered roads were remade and light railways laid down across the waste behind the front line. Preparations began for a renewed onslaught in the spring. With the rest, the companies of the Heavy Section were to be very busy that winter. The increasing difficulties of the attack, and their own demonstration of how those difficulties might be overcome, had established their position as a valuable, if not yet an indispensable, auxiliary. But they had fought their last battle as individual units. The corps was growing; and they withdrew from the line only to form the nucleus of the four battalions which were to be raised in France.

IV.

An immense amount of nonsense was talked and written about the first appearance of the tanks in the field. The newspapers, as perhaps was to be expected, excelled themselves in absurdity. Any lie was good enough for them. The war correspondents, in their safe and comfortable château in the valley of the Ternoise, ransacked Scripture and zoology for suitable similes. There is a story of those halfpenny strategists, Messrs Gibbs and Beach Thomas,[7] tossing up a coin (no doubt a halfpenny) to decide who should use the word Behemoth. All this ignorant flapdoodle did an ill service to the new unit. Its exploits in battle had been witnessed by few; but tens of thousands had seen tanks lying derelict in the Somme mud or holding up traffic through mechanical breakdowns on important thoroughfares; and exaggerated encomiums in ill-informed newspaper articles only brought ridicule upon men who had done wonders in circumstances inconceivable to the writers. The Heavy Section, in its early days, is said to have been the victim of more chaff (and not always good-natured) than any unit except the London Scottish. People at home, who believe anything they read in print and have the Press they deserve, naturally expected miracles of machines which pushed down houses on battlefields, where, in fact, no house had been standing for months.

No one disliked this sort of thing more than the men concerned. They knew well enough where they had succeeded and where they had failed. The necessary experience had been bought; and there were many lessons to be learnt from it and carried into practice, so far as was possible, before the next battle. It was proved that the Mark I. was designed on sound lines, and needed only a few structural alterations to make it equal to all reasonable demands until such time as it could be superseded by an altogether superior type. The “Hydraulic Stabiliser” or tail, for example, must be got rid of. Its essential worthlessness was discovered quite early, when a tank came under shell-fire while waiting for zero behind our front line. The whole crew being inside for shelter, with the engine just ticking over, a projectile burst under the stern and blew the tail away without doing any other damage: amid the general uproar no one was aware of what had happened; and the tank moved forward and was in action for some time before the loss of this absurd appendage was discovered. Other faults to be remedied were the position of the petrol-tanks, the overhead exhaust-outlet, emitting conspicuous flame and sparks, the needlessly broad sponsons (accessories in many cases of ditching), and the lack of any gear by which a ditched tank could free itself under its own power. “Ditching,” an unpleasant commonplace of tank warfare, can happen in many ways; but in essentials this misfortune means one of two things. Either the machine has got itself into such a position (_e.g._, tail down in a wide and very deep trench) that although the tracks will grip the surface the engine-power is insufficient to pull it out; or the belly, with a dead weight of nearly 30 tons on top of it, is resting on comparatively hard ground, while the tracks, happening on soft patches, cut through without gripping at all. The tank, for example, is moving parallel and close to the forward edge of a trench. The edge gives way under the weight of the near track, which slides down until the sponson on that side hits the rear wall of the trench. The fallen track is now revolving uselessly in air or crumbling earth, the belly of the tank is resting heavily on the parapet, and the efforts of the outer track, still on firm ground, to overcome the consequent friction, only succeed in embedding the sponson in the parados until the engine can no longer force the machine forward. What was needed was something which would give one or both tracks a more secure grip or leverage in such cases. After the experiences on the Somme, two separate devices were adopted. For permanent use over soft ground iron shoes or “spuds” were clamped at intervals along the tracks; while in addition each tank carried a pair of what were known as torpedo-booms. These were cigar-shaped objects of wood and iron, about six feet long, to be clamped to each track by chains when the tank became ditched. As the tracks revolved, these booms in theory were pulled round underneath the tank until they became jammed in the ground, when, the tracks being thus brought to a standstill, the machine was able to climb out over them in the normal way. In practice, although of great use on many occasions, these booms were found to be too small and weak. In soft ground they often were dragged under the whole length of the tank and up to the surface again behind: in hard ground, the chains were apt to part under the strain. A greatly superior development of this idea (to be described in its turn) was tested and adopted after the battle of Messines.

I offer no apology for this digression on unditching. Throughout the battles of the Somme, Arras, Messines, and Ypres, by day and by night, during approach-marches and under fire in action, this unsavoury process might fall at any moment to the lot of any crew. On at least one occasion at Ypres a whole company of twelve tanks was ditched hopelessly before it even reached the front line, and a few days before, in the same dismal area, out of two companies of the same battalion (my own) which went over at zero, only one tank returned before the end of the day. One other had been blown to pieces, the remaining twenty-one were ditched, temporarily or for ever, here and there about our little sector of that water-logged battlefield. Most of them had struggled somehow into action, and some were extricated from bogs or shell-holes half a dozen times before their exhausted crews gave up the struggle. At least another fifty from other battalions were in similar case that evening. Ypres, of course, was exceptional, and tanks should never have been used there; but until their proper function was discovered at Cambrai, he was a fortunate commander who could take his machine in and out of action anywhere without a case of ditching on the way.

Reverting finally to the lessons of the Somme, it was felt by most officers and men of the new unit that, well as they had done on many occasions, more might have been accomplished. They suffered none of the delusions which the Press was fostering at home. The lack of foresight in some directions was obvious enough now. At the same time it was realised that the tanks had been put to an unreasonable test. They were constructed for fair ground and fair weather; they had been granted neither. There had been insufficient time to train the crews. On account of the secrecy maintained at the start, it was impossible for infantry commanders and their men to become acquainted with tank tactics; but this want of co-operation was not remedied, as it should have been, before the later actions. Springing from the same source of ignorance was the mischievous tendency to use tanks in driblets instead of in masses--a tendency which the Tank Corps had to fight throughout its existence.

It was proved that the preparations for a battle required an infinite amount of work and forethought. There were contingencies to be considered such as affected no other arm. Especially was this the case with reconnaissance; for success or failure hung upon the tank commanders’ familiarity, from thorough coaching beforehand, with the ground they had to cover. Once in action, they could see very little, while maps were almost useless in a devastated country, and each officer instead had to carry in his mind the look of almost every tree-stump and every mound of rubble.

If tanks lost their way, apart from other consequences, the infantry were very apt to follow them and lose their own direction. Another difficulty, springing in part from the small size of the unit and in part from the bad practice of employing tanks in driblets, was that subaltern tank officers often found themselves working alone or in pairs with divisions or even with corps, and conferring on tactical points with G.S.O.I.’s and Generals. They had not the experience or authority needed to combat ill-advised demands, and as a result were often set to perform impossible tasks.

If the Heavy Section was to remain, still more if it was to be expanded, everything went to show the urgent need of reforming it as a fully equipped and independent organisation like the Flying Corps, whose requirements in many ways resembled its own. It must have its responsible staff to decide finally all questions dealing purely with tank warfare, its intelligence branch run on individual lines, its technical officers, its signal service, its workshops and depots, its transport and supply. Such a proposal was put forward after the first Somme battles, and as a proof that the mistakes and disappointments of that period were outweighed greatly on the credit side, the suggestion was adopted at once. Before the end of September the expansion and new establishment were authorised. Lieut.-Colonel Elles was given the command in France; and the building of a thousand improved tanks was put in hand without delay.