Alam Halfa and Alamein
CHAPTER 8 — Withdrawal of the Panzer Army
Withdrawal of the Panzer Army
SEPTEMBER opened with the Panzer Army lined up across the south of Alam el Halfa in a good position to commence the assault on Eighth Army's rear, as envisaged in Rommel's plan, but exactly twenty-four hours behind schedule. The element of surprise had thus been lost but no major regrouping had been made by the British, so that an opinion was growing among the Panzer Army staff that the Eighth Army was short of reserves and unlikely to attempt a large-scale counter-attack. In spite of the delay, the opportunity was in fact still waiting for the Panzer Army to give the British the ‘pretty thorough beating’ of Rommel's secondary hope.1
In detail, the two panzer divisions were well poised to give 22 Armoured Brigade some fairly rough handling, even to the extent of eliminating it as a fighting force, for 23 Armoured Brigade's Valentines could have offered only limited assistance, while 8 Armoured Brigade's Grants were cut off by 15 Panzer Division and were so far away that they could have been dealt with separately later. The line of communication from the spearhead to the Qattara Box was being strengthened hourly and the stage was all set for the Panzer Army to open its assault.
1 Rommel Papers, p. 275.
Dawn on 1 September brought news which added to Rommel's misgivings. First, the Reconnaissance Group sent in a report that it was scarcely battle-worthy. In the approach march its vehicles had suffered severely from ground and air action, mines, and mechanical trouble. In its laager overnight near Samaket Gaballa, it had been caught in the light of Air Force flares and pattern bombed. Now about a third of its vehicles was completely destroyed or in need of repair and it had a large number of casualties to be evacuated to the rear.
After this Rommel learnt that a strong force of tanks with infantry in support was attacking 164 Division in the coastal sector, while 15 Division reported a force of British tanks approaching from the east.
Meanwhile, 15 Division, renewing its attempt to take Point 132, continued its complaints of lack of co-operation from 21 Panzer Division, whose commander countered by questioning the accuracy of 15 Division's reports. This led the Corps Commander himself to drive out on a reconnaissance in which he discovered that the two divisions were separated by a much wider expanse of unoccupied desert than their situation reports indicated. His arrival at 15 Division caused further confusion for, according to messages sent out by him, he accepted the accuracy of the division's obviously inaccurate map-reading of its own position and the position of Point 132. He therefore ordered 21 Division to send tanks at once to assist in the attack on the point, only to receive the curious reply that the latter division's tanks were needed ‘for special duties’. The only explanation of this reply is that the divisional commander feared to give over the radio his true situation, which was that his tanks were out of petrol and that he did not dare to manoeuvre his vehicles more than necessary until his supply columns, ‘expected hourly’,1 put in an appearance. The division was in fact immobilised through lack of petrol until the middle of the afternoon, when a limited amount was brought to it.
1 Diary of 21 Pz Div, 1 Sep 1942. GMDS 27641/5.
While the tanks were skirmishing on the south of Alam el Halfa, there occurred ‘one of those little actions customarily described … as a “bad show”’,1 the Australian diversion code-named bulimba. This operation had originally been planned over a week earlier as one in the sequence of major raids for harassing and information ordered by Montgomery. The Australian contribution, however, had somehow developed on a greater scale and with a somewhat different conception from the raids already laid on by the South Africans and New Zealanders. According to its operation order, bulimba was intended as an ‘immediate counter-stroke to enemy attack or as a diversion prior to enemy attack’. Set in motion some thirty-six hours after the enemy had attacked, its immediacy as a counter-stroke could be questioned. That it was intended as ‘a diversion prior to enemy attack’, that is to harass and deter a possible enemy assault in this sector to link up with the spearhead, seems unlikely. It appears to have been allowed to proceed purely in the hope of either diverting forces from the enemy's main advance or preventing the Panzer Army troops in defence from being used to reinforce the striking force. However, reports and despatches remain vague as to the operation's exact purpose.
In short, an infantry force was to push forward a salient into the enemy's lines, clear it, and hold it as a firm base from which tanks would advance to ‘exploit’—a military term which meant many things to many men. The exploitation was to cease about 3 p.m., when the tanks would retire through the firm base, to be followed after dark by the infantry. No material gain of ground was intended.
The infantry, four companies of Queenslanders of 2/15 Infantry Battalion, crossed the start line on time. As they approached the area chosen to be occupied as the firm base, heavy artillery concentrations kept the enemy in cover, but as soon as the shelling lifted to its next targets, the enemy came vigorously to life with automatics and mortars. Most of the Australians reached their objectives but they had struck an area held by German troops, who resisted tenaciously and mounted immediate local counter-attacks whenever they were driven from their positions. For some time communications broke down, battalion headquarters losing touch both with its companies and with the squadron of Valentines of 40 Royal Tank Regiment which was to come up in support and start the exploiting.
The tanks themselves experienced difficulty in negotiating the narrow gaps cut in haste by the engineers through the British and enemy minefields. Before they got through the enemy mines, the Valentines came under fire and, with several of their number ablaze and others immobilised by mines or gunfire, halted well short of the infantry.
The second-in-command of the Queensland battalion, Major Grace, taking charge when his commanding officer was wounded, decided that the uncertain communications and lack of close co-ordination between tanks and infantry made the task of securing a firm base practically impossible and the likelihood of the tanks going on to exploit remote. Using the one reliable means of communication, the wireless link between the artillery's forward observation officer and the guns, he called for covering fire from all available weapons and sent word by runners to the infantry for them to retire as soon as the fire commenced. Four hours after the operation started, the companies were on their way out.
It is difficult to assess whether this raid had any value or not. Rommel may have been influenced, if only to the extent of calling back some of his Africa Corps tanks to form a mobile counter-attack force in case a similar attack was repeated. It may, on the page 107 other hand, have proved to him that his infantry defences could stand unsupported by tanks, for no enemy tanks took part in the defence.
A tally of casualties showed that the Australians lost 15 men killed and 120 wounded, and nine tanks were knocked out. Against these losses, the enemy lost probably 150 killed, while the Australian infantry maintained their reputation by bringing out of the disorder and confusion nearly 140 prisoners, most of whom were Germans of 382 Infantry Regiment.
While this early morning local battle was being fought far away near the coast, another isolated and inconclusive skirmish by British forces had begun off the south of Alam el Halfa. On orders to make contact with 22 Armoured Brigade, the regiments of 8 Armoured Brigade had set off to work their way westwards across the southern face of the ridge defences. The advance was cautious as the brigade commander had been told not to become too deeply involved or risk unnecessary casualties through lack of experience and the incomplete state of training in his brigade, a state that became obvious soon after the start, when one regiment ran out of petrol. It was not until about 8.30 a.m. that the leading tanks, then off to the south-east of the el Halfa ridge, came into contact with 15 Panzer Division, whose eastern flank was covered with a gun line in which were several 7·62-millimetre self-propelled guns of Russian origin. Under fire from these guns, which the brigade thought were the dreaded 88s, the tanks halted while the brigadier made arrangements with 44 Division for covering fire under which he intended to lead his brigade northwards and thence, in the shelter of the infantry minefields, on to the west. No sooner had this movement started than a heavy Stuka raid came in on the area between the two el Halfa boxes, where the headquarters of 13 Corps, 10 Armoured Division and 44 Division were all situated. The resulting disruption of communications upset the co-ordination between guns and tanks and prevented the leading regiment finding a gap in the minefields through which it could enter the infantry defences. Meanwhile 15 Panzer Division's tanks moved up and forced the brigade to withdraw to the east. A later repetition of the same manoeuvre was no more successful so that, by the end of the day, 15 Panzer Division was still sitting squarely between Eighth Army's two heavy armoured brigades. Thirteen Grants and three Crusaders of 8 Armoured Brigade were lost by enemy fire but the casualties in men were relatively light—three killed and twenty-six wounded.
For the New Zealanders in their box, this second day of the battle was hardly more eventful than the first. Artillery observers who had gone out before dawn to the high ground overlooking Deir el Muhafid found themselves faced by a number of Italian tanks and retired hurriedly under covering fire from the Stuart tanks of a Divisional Cavalry patrol. The Italians showed little inclination to do more than some long-range sniping at the Cavalry Stuarts and, when a patrol of Valentines from 23 Armoured Brigade appeared, they withdrew out of range. This permitted the observers to return and watch, in the middle of the afternoon, a large force of both German and Italian tanks with infantry in trucks drive across the east of Muhafid and turn north-east towards 23 Armoured Brigade's position. The Valentines and Stuarts opened up on this target, to be joined by all available field guns. Numerous hits were claimed on the vehicles but, though the enemy did not seem to be advancing with any great determination, by nightfall a few tanks and infantry were little more than a mile from the south-east corner of the New Zealand Box. The movement gave rise to a fear that the enemy might be preparing to mount an attack at dawn next morning, so General Freyberg cancelled plans for a repetition of the previous night's raids by tanks and infantry and instead ordered listening posts to be set out and maintained until dawn.
The enemy's movement was in fact not a concerted operation but a number of only loosely related activities. The tanks were mainly those of Littorio Division, with no immediate aggressive task other than to protect 21 Panzer Division's left flank and link the spearhead with the infantry covering the line of communication rearward. To Littorio's left, that is, somewhere south of 23 and 28 Battalions, troops of 90 Light Division with a few tanks in support were attempting their original task of forming a line across the south of the British defences in preparation for a possible attack to the north in conjunction with the armoured spearhead.
Two prisoners of 90 Light Division who were gathered in by a 23 Battalion patrol just after dark told a story of a hard day. Two parties, of about thirty-six men altogether, had set off that morning from the vicinity of Deir el Munassib with orders to advance to the north-east. Under heavy artillery fire the men had halted until they were joined by six tanks. The tanks then led off again but, after one had been knocked out, the rest retired, leaving the infantry under fire in the open. By evening all but three of the men were casualties. After dusk, these three set off, one openly stating his intention to desert and going off on his own. The other page 109 two, with apparently only a vague idea of direction, must have wandered in a northerly direction until they fell into the hands of the New Zealand patrol. Their story illustrated the typical probing method of advance used by the Germans, as well as the effectiveness of the British defensive fire, especially that of the 25-pounders.
The second day of the battle, which saw the Australian raid and some minor skirmishes on the Alam el Halfa front, ended with the bulk of the two armies in much the same positions as they had occupied at dawn. The three British armoured brigades had already lost some fifty tanks, more than half of them Grants, in the inconclusive skirmishing, apart from losses sustained by the columns of 7 Armoured Division. The Panzer Army's tank strength was down by about ninety German tanks and an unknown number of Italian, but many of these had suffered mechanical breakdown or minor damage from shelling and bombing and were repairable. The total German and Italian losses by actual encounter appear to have been no greater, and may have been less, than the British losses by the same cause, while the loss of so many of the Grants made it plain that the British armour was still no match for the German panzer divisions.
By the morning of 1 September the Eighth Army had garnered sufficient information to account for all the major formations of the Panzer Army. Patrols, air reconnaissance and other sources confirmed that there was no ominous gathering of forces along the static front, so that General Montgomery could confidently assume that the drive on Alam el Halfa was the only threat to be dealt with, at least in the immediate future.
Unaware that the stringency of supplies had already brought Rommel's offensive to a stop, Montgomery was concerned about the defences between Alam el Halfa and the sea. He therefore arranged to combine the strengthening of these defences with the gathering of his reserves. One brigade of the newly arrived 51 (Highland) Division was brought from its training and acclimatisation in the Delta defences to take the place of 151 Brigade in the Amiriya box, the latter formation moving up to reinforce Localities G and H, two prepared boxes which linked the north-east corner of the Alam el Halfa defences to the sea. The South African Division was ordered to take responsibility for Localities C and D on the Gebel Bein Gabir in order to release the troops of 5 Indian Infantry Brigade to reinforce the New Zealand Box. By the end of this rearrangement the New Zealand Division would be in command of four infantry brigades, two to man the western face of the box and two for the proposed pincer movement to Himeimat. Though this ‘general post’ of brigades entailed numerous and complicated exchanges of field and medium artillery and other supporting arms, Montgomery wanted it complete early on 2 September so that, if necessary, operations could commence that evening. In the event it took longer than expected, some of the delay being caused by the commander of the South Africans, Major-General D. H. Pienaar, who objected to the detachment of one of his brigades both on principle and on the practical grounds that his sector was too large to be held by two brigades alone. His objections were overruled by Montgomery, who directed that 5 Indian Division should extend its front to take over part of the South African sector.page 111
Towards the evening of 1 September Freyberg received a message from 13 Corps with a decision on the methods already discussed earlier for the attack to the south. The first stage proposed a limited advance of some two to three miles to gain the northern sides of the Munassib and Muhafid depressions. This was to be followed, according to the degree of success of the first stage and of the enemy reaction, by a further advance through the depressions and beyond. Apart from a proviso in the corps' order that the Indian brigade should ‘eventually’ be sited in the north of the New Zealand Box, where by a simple extension of the boundary it could be taken back under command by 5 Indian Division, Freyberg understood he could use any of the four brigades under his command for the coming operation, the first stage of which was to begin on the night of 2 September, that is in twenty-four hours' time.
But while Montgomery was planning slowly and cautiously, the opportunity to strike really hard at the Panzer Army was passing. About the time that the corps' order reached Freyberg, Rommel had discounted the sanguine expectations of 15 Panzer Division and had made his final decision to withdraw. Only the arrival of almost unlimited quantities of petrol in the front line could have persuaded him to change his mind. The long advance and heavy going of the first night had so drained his reserves that supply could not keep level with the demand, even though his army had spent two days of relative inaction. Kesselring's air lift had proved a broken reed while British air and sea action had taken such toll of shipping and land transport that only extreme efforts could get enough petrol up to the Panzer Army to permit it to withdraw, let alone manoeuvre in an attack.
At this point, when the Panzer Army has been taken to the limit of its advance, it is worth considering the much publicised story of the ‘going map’ which was supposed to have deceived Rommel into choosing the worst possible route that led him into a trap of soft sand. A New Zealand officer who was concerned in the actual preparation and printing of the map has given his story, and various other versions have appeared in print telling how the idea was conceived, and how the fake map was printed in great secrecy, marked and stained to give the appearance of use, and finally left in an abandoned vehicle for the enemy to find. Evidence that the map was found and used by Rommel's staff in their planning was given by General von Thoma on direct questioning after his capture later in the year, but this officer did not arrive in Africa until after the Alam el Halfa battle was well over. Whether the map was used or not, the effect on operations of the false page 112 information contained in it must have been negligible. The faked going details commenced east of the point that the Panzer Army reached on the morning of 31 August. On Rommel's original planning, the advance should then have continued in an easterly direction for some distance before turning north. As it was, the failure to keep to the time schedule caused Rommel to order an advance from this point in a north-easterly direction and, although Africa Corps met some bad going, Rommel's choice of this route was plainly influenced by the circumstances and not by any map. Had the advance gone according to plan, some of the Panzer Army might have been drawn in to areas of bad going, but it should not be forgotten that, in the Panzer Army's tactics, the planning of routes across the desert had been simplified to the point where subordinate commanders were given objectives and expected to reach them. Difficulties of terrain were treated as forms of opposition, in the same way as minefields or points of resistance, to be discovered by quick reconnaisance and either overcome or avoided. The first night's crossing of the southern minefields illustrates these tactics, for it is plain that the Panzer Army made the most cursory of reconnaissances and then proceeded to feel its way across. More than anything else, the importance laid on the false going map indicates that not only those responsible for the ruse but those who later extolled its improbable success still had only a hazy idea of German tactical methods.
New Zealand patrols brought back news of considerable activity in the two depressions, Angar and Munassib, where the Axis troops were obviously digging positions of some strength. Between the depressions and the box there were numerous groups of trenches, fully or partially dug but unoccupied. Several of the slit trenches were found to contain booby traps, which were disarmed by engineers with the patrols. No one knew whether the Axis troops intended to occupy these positions when they were completed, or whether they had abandoned them as too close to New Zealand observation.
With Africa Corps patrol reports offering no hint of British night activity between Alam el Halfa and the New Zealand Box, followed by a quiet dawn, Rommel was convinced that he did not have to fear a counter-attack for some time, his opinion being that the British were too short of reserves to risk moving from their defences. Montgomery at about the same time gave out his opinion that the Axis forces, probably through shortages of supply, were adopting a defensive attitude, at least for the time being. Both commanders therefore felt free to go ahead with their own plans without undue interference, Montgomery with the provision of his counter-attack forces, Rommel with a withdrawal before the ‘slow reaction of the British command’ could hinder him.1
Rommel prefaced his withdrawal orders with the explanation that British air supremacy and the sinking of several tankers made a continuation of the offensive impossible. Accordingly the Panzer Army was to retire in bounds to positions behind the British minefields that ran north from Himeimat, the new line to be linked to the old through Deir el Angar and the Qattara Box. Before the general withdrawal commenced, both panzer divisions were to supply detachments of tanks and infantry as mobile reserves to reinforce any part of the front against which the British might attack. Rommel was particularly sensitive to the Qattara Box area, where his new line hinged to the static front.
Observers in Eighth Army, though unaware that a major withdrawal was commencing, were quick to note that outlying columns and patrols were being pulled in on the morning of 2 September. On renewing its advance at dawn 8 Armoured Brigade passed across the south of the Alam el Halfa box unmolested, to join 22 Armoured Brigade about 9.30 a.m. Columns of 7 Armoured Division discovered that positions in which 15 Panzer Division's troops had laagered overnight had been vacated but, when they tried to follow up, were met by the gun line of the rearguard. Well to the south a column of 4 Hussars nipped off an isolated party of the Reconnaissance Group, shooting up several trucks and taking some prisoners.
By the middle of the morning the field guns of 10 Armoured and 44 Divisions were falling silent as few targets remained within range. The armoured division then took some of its batteries from their dug-in positions and set them up further to the south. By midday a thick dust-storm was raging, hindering observation both from the ground and the air. Several New Zealand artillery observation officers took advantage of the dust to settle themselves in better vantage points overlooking the line of the depressions on the south of the box and, when the dust began to settle about two o'clock, reported a wealth of targets as numerous enemy columns, also taking advantage of the low visibility, were moving to their positions for the withdrawal. On orders from Eighth Army for maximum harassing fire, all British guns within range maintained constant fire until dusk, several batteries expending over 1000 rounds each during the afternoon and evening.
A curious incident occurred this day in the sector north of the New Zealand Box. English troops of the Essex Regiment on Ruweisat Ridge were approached by a German officer and NCO carrying a white flag. The officer, identified later as a member of the Ramcke Brigade, demanded immediate surrender on the grounds that Rommel had surrounded the Eighth Army. The two optimists were escorted back to 5 Indian Division's headquarters and eventually on to the headquarters of 30 Corps. After interrogation, it is understood, they were finally returned to their own lines. page 115 Some days later Rommel issued an order forbidding keen but misguided officers from acting as parlementaires without official approval.
Early on 2 September General Freyberg learnt with some relief that the operation planned for the coming night was to be postponed for twenty-four hours. The movement of 5 Indian Infantry Brigade had been delayed and its troops only began to arrive in the box in the late afternoon. The first battalion to appear took over the sector held by 26 Battalion, which then moved back to an uncomfortable bivouac in the rear of the sector. During the evening and night the remainder of the Indian brigade relieved 132 Brigade, whose troops moved back to bivouac among the gun lines in the northern part of the box. As these reliefs settled in, the two New Zealand brigades sent out patrols to the south-west and south to reconnoitre the ground and enemy defences in the area over which the proposed operation would occur.
One patrol, under Second-Lieutenant Mowat1 of 25 Battalion, had a brisk engagement on the edge of Deir el Angar and had to be assisted in disengaging by covering fire from field guns and mortars. No other patrol encountered the enemy, but the total of the information brought back was that the main enemy line had been prepared along the northern edges of the depressions Angar, Alinda, Munassib, and Muhafid. North of this the section and platoon positions discovered the previous night were still unoccupied though some showed signs of further work. More booby traps were found and disarmed. These partially prepared positions reached to within a mile of the southern minefield of the New Zealand Box.
While these patrols were out, the British guns, on Montgomery's instructions, harassed the enemy constantly throughout the hours of darkness, while seventy-two sorties were flown by the Air Force between dusk and dawn. Two 4000 lb bombs were dropped this night, both causing large fires to show that they had found targets of some kind. According to the German records, 300 aircraft dropped 2400 bombs and caused such damage to transport that some units were nearly immobilised.