i

THE Panzer Army of Africa was not accustomed to attacking in darkness and Rommel did not in fact plan the Alam Halfa encirclement as a night operation. His intention was to concentrate his forces as soon as the evening shadows made them less vulnerable to air observation and attack, to lead them through what both he and his staff considered were minor obstacles of the minefields and patrols that spread south of the main front, and have his spearhead formed up by dawn facing north against the Eighth Army's rear defences. From the concentration areas to the dawn positions meant a journey of up to 40 miles over broken country, and for this sufficient local visibility was necessary to keep the columns of vehicles in contact and control.

As each night passed after the moon reached the full, the period of darkness between sunset and moonrise was increasing while the hours of moonlight and the brightness of the moon both decreased. By 30 August the moon was appearing about three hours after sunset, so that the Panzer Army, setting off after daylight had completely faded, had to start its trek in darkness in which both contact and control were difficult. There was then a limited period of moonlight in which any confusion could be sorted out and the columns could reach their destinations before daylight revealed their positions and intentions. The planners on Rommel's staff held that the night of 30–31 August was the last on which the operation stood a chance of success.

To postpone the operation until the September full moon was, in Rommel's opinion, to give the battle away to the Eighth Army, for by this time he knew he could not rely on any large addition of German troops to offset the arrival of men and equipment known to be flowing in to the British base in Egypt. It was, in fact, now or never.

At a meeting on 27 August, Rommel discussed his Panzer Army's needs and chances with Kesselring and Cavallero. The Italian marshal was obviously infected with Rommel's enthusiasm, if Ciano's diary can be accepted as evidence, but Kesselring's opinion is difficult to unravel from comments after the event. Both of them, however, gave sufficiently earnest promises that the one real shortage, of petrol, would be overcome within forty-eight hours. On the basis of these promises, Rommel set the date of the offensive tentatively for the 29th.

It seems strange that Rommel had not grasped that the Italian commander, in spite of his lip service, was incapable of galvanising the Italian supply organisation into action, and that Kesselring could not or would not use his influence with the German High Command to obtain the support requested. Why he accepted their promises can only be attributed to a combination of his own enthusiasm and his feeling that the offensive must be ‘now or never’. Although he has been accused of ignorance of logistics, it would be fairer to charge him with a lack of appreciation of the human element in the problem.

Kesselring's attitude to Rommel and his plans is enigmatic. Through his position and rank he should have been able to exert considerable influence on the planning and supply for the African campaign. Yet in the records and reminiscences of this period, he emerges in a negative light. He does not appear to have encouraged Rommel, at least very positively, or to have acted as a restraining influence of any strength. Though his role of Commander-in-Chief South placed him theoretically above Rommel, the latter's personal repute and favour with Hitler may have made his position difficult. He was clearly the superior in his own sphere, the command of the Luftwaffe, though his control over the Italian air force was limited by the Italians' intransigence.

The explanation of Kesselring's equivocal attitude may be that, while Rommel was Hitler's man, Kesselring was the representative of the High Command with the task of reporting and, to some extent, controlling the campaign. With success still in the air he might not care to be too critical of Rommel, especially as his comments could reach Hitler's ears, but he plainly had a wider vision than Rommel of the relative position of the African campaign in the broader strategy. From the little that emerges of his shadowy figure at this period, he might be accused of playing safe so that, should Rommel succeed, he would share the glory but, if Rommel failed, he could avoid the blame.

Much of the story of the two men is told in the letters Rommel wrote home to his wife. When he was able to take his mind off the battle at the end of July to consider why his German requirements were not being met, Rommel initially chose to put the blame on Italian self-interest rather than inefficiency and thought the solution lay with the liaison official in Rome, von Rintelen, who ‘lets himself be done in the eye, for the Italian supplies are working excellently’.¹ After a visit by Kesselring to Panzer Army headquarters on 9 August, when the problems of the coming offensive and its supply were discussed, he received such encouragement that he wrote, ‘We reached agreement over what is to happen. Now it's a question of making full use of the few weeks to get ready. The situation is changing daily to my advantage.’²

Arising from this same meeting, Rommel sent a proposal to the German High Command that Kesselring should be given special powers to control Mediterranean shipping, on the grounds that

Kesselring had a personal interest in helping us at Alamein; he had considerable strength of will, a first-class talent for diplomacy and organisation, and a considerable knowledge of technical matters.

Kesselring had the Luftwaffe and Goering behind him and could thus command sufficient support at the highest level to enable him to tackle questions of high policy in relation to Italy.³

Although, according to Rommel, his suggestion was not acted on early enough or in the form he wanted, Kesselring did in fact make a show of exerting his influence, but only to the effect that, just before the meeting on 27 August, Rommel wrote home:

Kesselring is coming today for a long talk over the most acute of our problems. He, too, often has a tough job in Rome. He gets plenty of promises, but few are kept. His over-optimism concerning these blighters has brought him bitter disappointments.⁴

Rommel had the impatience of the enthusiast so that, in the realms of supply over which he had no direct control, he was willing to accept promises at their face value. His own army, by restricting its activities and hoarding supplies, was rested, reorganised, and sufficiently equipped with almost everything except petrol for a short, but possibly decisive, action. Much was made for British propaganda purposes of Rommel's stated intention to reach the Nile, set against his story after the battle that the operation was just a ‘reconnaissance in force’. But Rommel's true attitude appears in a letter he wrote his wife a few hours before he sent his striking force on its way:

Many of my worries have been by no means satisfactorily settled and we have some very grave shortages. But I've taken the risk, for it will be a long time before we get such favourable conditions of moonlight, relative strengths, etc., again…. If our blow succeeds, it might go some way towards deciding the whole course of the war. If it fails, at least I hope to give the enemy a pretty thorough beating….⁴

The following night, 28th-29th, the assembly was continued, but it soon became obvious that the original five-night plan could not successfully be condensed into two nights. It was with considerable relief, therefore, that the army learnt from Rommel on the morning of 29 August that the offensive would not start until the evening of the 30th.

Rommel's decision, though possibly influenced by the delays encountered, was primarily based on news he received shortly after the meeting, that three ships carrying petrol had been sunk or damaged. The vehicles of his mobile formations had been ‘topped up’ by the expedient of draining the forward dumps of practically every gallon, and it was only after he had been assured again by Kesselring and Cavallero that more tankers were on the way and that the Luftwaffe would assist in the transport from the rear to the forward dumps, that Rommel gave his final decision. Even then, the offensive might have been postponed again had not news reached the Panzer Army headquarters on the 30th that a tanker had arrived safely in Tobruk and that the air lift had commenced.

Throughout the 30th, the Luftwaffe flew constant patrols over the southern part of the front to drive off British reconnaissance planes while the German and Italian formations sorted themselves out after their hurried assembly. With a last-minute distribution of petrol which had just been brought up, the two German panzer divisions recorded that they had enough fuel to take their tanks about 100 miles and their other vehicles about 150.

The start line from which the German-Italian offensive was to drive to the east ran practically due south from the Qattara Box. On the far right flank, that is, on the Taqa Plateau just north of the edge of the Qattara Depression, the Reconnaissance Group formed up. This force comprised 3, 33, and 580 Reconnaissance Units of Africa Corps and a composite reconnaissance group from the Italian 20 Corps, the whole under the command of the headquarters of 15 Rifle Brigade of 90 Light Division. It appears to have had about fifty armoured cars, some light tanks, and a considerable number of other vehicles including tracked troop-carriers and mobile guns, and was about 2000 men strong.

Along with the Reconnaissance Group there was a battalion of Italian parachutists from Folgore Division which had the limited task of occupying Qaret el Himeimat, an isolated hill of some 216 metres above sea level that gave a commanding view over a large expanse of surrounding desert.

In the area around Gebel Sanhur immediately to the north of Taqa the main body of Africa Corps lined up with 15 Panzer Division on the right and 21 Panzer Division on the left. These two divisions commanded most of the German tanks, and each division with its motorised infantry regiment about 2500 strong, tank crews, engineers, artillery, anti-tank and other units, mustered slightly over 6000 all ranks.

Still further north, slightly to the left rear of Africa Corps, the Italian 20 Armoured Corps congregated on the El Kharita plain. The corps reached the start line with fewer than 250 of the lightly armoured and mechanically unreliable tanks of Ariete and Littorio Armoured Divisions. Its infantry component was the Trieste Motorised Division. As much of the Corps' artillery and about six infantry battalions for which there was no transport remained behind, the effective strength of the Italian mobile force is difficult to assess, but it was probably about half 20 Corps' paper strength of some 16,000 men.

The 90th Light Division, with a strength of just under 4000, formed up immediately south of the Qattara Box with, on its left, a mixed force made up of two German battalions from Ramcke Brigade, two Italian battalions from both Folgore and Brescia divisions and a group of the artillery and men from the infantry battalions of 20 Corps without transport, the whole under the nominal command of 10 Italian Corps.

The design of the advance was simple. Moving out of their assembly areas at dusk, the armoured and motorised formations were to form up and cross their start lines at ten o'clock of the evening on a course slightly south of east. After passing the New Zealand Box, they were to swing first due east and then north-east. On the rim of this wheel, the Reconnaissance Group was allowed five and a half hours by the plan to complete a journey of about 40 miles to reach its dawn objective. At the hub, the Italian Ariete Division, with only 20 miles to go, was allowed five hours. By 5 a.m. the whole striking force was expected to be lined up facing north along a front of some 15 miles, with the left flank resting on the depression known as Deir el Tarfa. From this front the advance was to drive due north, the Reconnaissance Group with another long journey to cut the coastal road and railway between El Hammam and El Imayid and protect the eastern flank, while Africa Corps, 15 Panzer Division still on the right and 21 Panzer Division on the left, crossed the Alam el Halfa and Gebel Bein Gabir ridges, and 20 Corps, with Littorio right and Ariete left, moved on to the western end of Alam el Halfa. In order to guard the lines of communication, 90 Light Division was page 89 to move inside the wheel to the Deir el Muhafid area and link up with the left of 20 Corps, and the mixed force under 10 Italian Corps was to fill the gap between 90 Light and the Qattara Box. The final result would have had the Eighth Army bottled up against the sea with strong forces on all three landward sides.

sketch taken from a map on which rommel marked his intended movements towards cairo and the suez canal should he break through at alamein

Had the initial plan succeeded and brought the British forces to a sufficient state of confusion as at Gazala, a second phase was to be attempted. Leaving any small centres of resistance that remained to be dealt with by the non-mobile troops, 90 Light and 15 Panzer divisions were to drive south-east direct for Cairo, the light division then making a dash for Suez. The Italian 20 Corps was to cut the Cairo-Alexandria road, 21 Panzer Division was to encircle Alexandria and despatch a column to Port Said, and the remainder of the Panzer Army was to invest Alexandria from the west.

It was an ambitious plan but Rommel was probably justified in attempting it. The disparity in strength was not so widely dissimilar from that prevailing at Gazala, and Rommel hoped to iron out any inequalities by the efficient handling of his armour against the slow reaction of the British commanders. It was, in fact, upon the work of the two German panzer divisions that the whole plan depended. It is interesting to note that the original plan held no provision for a channel to be cut through the main British defences, for example, along the Ruweisat Ridge.

Much has been made of the fact that, in comparison with the situation at Gazala, the British positions were more compact and closely defended and that the channel for the striking force was more confined, and comprised much heavily mined, broken, and difficult ground. These two factors naturally had a bearing on the outcome of the battle, but it should be remembered that, without undue loss or much delay, Rommel's striking force reached an area where the typical manoeuvring of the panzer units was possible. Why such manoeuvres were curtailed will appear as the story unfolds.

ii

The general dispositions of Eighth Army had not been changed when the Panzer Army advance commenced. The fixed defences from the coast to Alam Nayil were held by 9 Australian, 1 South African, 5 Indian and 2 New Zealand Divisions, in that order. The minefields running south from Alam Nayil to the Qattara Depression were patrolled by the columns of 7 Armoured Division, with 7 Motor Brigade responsible for the northern half and 4 Light Armoured Brigade the southern half of the sector. The motor brigade commanded five main groups, the 10 Royal Hussars Group with 41 Crusader tanks, the King's Dragoon Guards in three squadrons with a total of 57 armoured cars, and three mobile columns, 2 and 7 Rifle Brigade and 2 King's Royal Rifle Corps groups, page 91 each of approximately a battalion of motorised infantry supported by a battery of field guns, a troop each of anti-tank and anti-aircraft guns, and a platoon of machine guns. The light armoured brigade was divided into six main groups, of which 1 King's Royal Rifle Corps and 3 Royal Horse Artillery were mobile infantry and gun columns. The other four consisted of 11 Hussars and 12 Royal Lancers, each manning 54 armoured cars, 4 Hussars equipped with 50 Stuart tanks, and 3 County of London Yeomanry (3 Sharpshooters) with 27 Crusaders.

On paper 7 Armoured Division appeared much stronger than it was. With a front of over 15 miles, it had insufficient infantry and guns to man a continuous line, so that its various groups occupied temporary bases inside the minefields through which patrols were sent out into no-man's land to watch and harass the enemy. The division's orders were to impose the maximum possible delay on any enemy advance without getting so involved that any of its columns should become surrounded or overwhelmed. In short, its role was to harass and run. In the rear of its sector there were two dummy tank brigades and a partly completed dummy infantry position, manned by a few troops whose task was to demonstrate against any light columns in the hope of turning them to the north against the forces disposed on and around Alam el Halfa.

The true rearward defences of Eighth Army consisted of the two brigade boxes of 44 Division on Alam ei Halfa ridge, 133 Brigade on the west and 131 Brigade on the east.¹ These positions were now well dug in, encircled with minefields, and supported by a field regiment in each brigade box and numerous anti-tank guns. The division could also call on the two field regiments of 22 Armoured Brigade, which lay immediately to the south-west of 133 Brigade. This brigade provided the core of the army's armoured striking force, with its 92 Grants, 34 Crusaders and 40 Stuarts divided into four regiments, each with two heavy squadrons of Grants and one squadron of the lighter tanks, manned mainly by experienced crews.

The open desert south of 131 Brigade was guarded by 8 Armoured Brigade which possessed 72 Grants and 12 Crusaders. Its three regiments were still undergoing training and the brigade had not been in action before as a formation. Both 22 and 8 Armoured Brigades were under the command of 10 Armoured Division, whose headquarters lay between the two boxes on Alam Nayil.

For reserves, there were in the Middle East a number of tank brigades in process of being equipped and trained, from which about one complete brigade might have been assembled quickly to reinforce Eighth Army. The reserves of infantry available would probably have been the equivalent in numbers of at least two divisions. Amongst those training or refitting were 50 Division, of two weak brigades, 4 New Zealand Brigade at Maadi, a large part of 51 (Highland) Division which had begun to arrive at Suez in the middle of August, as well as units of Free French, Greek and Indian troops. Many of these were garrisoning the Delta defence zones and the rear landing grounds but could have been called forward in an emergency.

iii

The reaction to the start of the Axis offensive by Eighth Army on the ground has already been described. In the air, the Allied Air Force had been prepared for some time to lay on heavy assaults, with plans closely co-ordinated with those of the ground troops through Montgomery's insistence on close liaison between the army and air headquarters.

Small signs during the preceding days, such as reports from all reconnaissance aircraft that enemy vehicles tended to move from north to south, the enemy's own air patrols against reconnaissance and general lack of aggressive air activity over the front, all went towards convincing the Desert Air Force command that its assistance might soon be needed. For several nights towards the end of the month Axis aircraft had stepped up their bombing of the forward landing grounds, especially in the vicinity of Alexandria, and on the evening of 30 August enemy bombers came over in force to bomb the principal landing grounds at Amiriya, Wadi Natrun and Burg el Arab. None of these raids caused any grave damage to the sandy airstrips or the well dispersed aircraft, or affected to any degree the activities of the Desert Air Force.

In the early part of the night of 30 August, RAF bombers on routine harassing had noted and bombed concentrations of vehicles on the El Kharita plain, where the Italian 20 Corps was assembling. By the time these particular bombers had returned to their bases and reported their activities, Air Headquarters had heard from the army that the Axis attack was on its way and all available night bombers were got ready to take to the air. By 3 a.m. searching aircraft had picked up three main enemy columns and kept them under observation by successions of flares to guide the bomber squadrons. From then until dawn, according to observers in the New Zealand Box, there was hardly a minute in which flares and bombs ceased to fall on the desert to the south.

The German records make it plain that much of the delay in the planned night march was caused by air action for, although material damage was probably not great for the weight of bombs dropped, the morale effect on men trying to break a way in unknown desert through minefields on a hostile front was considerable. Many vehicles and even tanks received minor damage, while one bomb landed close to the headquarters group of Africa Corps, wounding the corps commander, General Nehring, and causing several casualties among the staff.

Ground action on the southern front commenced as early as 10 p.m., when heavy shelling began to fall on various points along the outer or ‘first’ minefield. This shelling itself was unusual, and, when the small patrols which normally stayed overnight on the west of the minefields began to fall back to the main laagers of the various 7 Armoured Division columns with reports of large and aggressive enemy patrols, it was obvious that something was afoot. A good half hour before 13 Corps called its troops to the alert, 7 Armoured Division ordered its columns to instant readiness. Following their orders not to get too closely involved and cut off, the columns opened fire on any of the enemy who approached the minefield, until retaliation became too heavy, when they disengaged and withdrew. The Germans found this fire particularly accurate and were considerably delayed by it in their attempts to lift the mines. No sooner had the ground forces disengaged than the Air Force appeared in strength overhead. The German method of dealing with minefields was simple but effective. Each advance guard, led by engineers with infantry and tanks in close support, drove forward until a minefield was either recognised or suspected, or until one of the vehicles was blown up. The engineers then debussed and, providing their own immediate protection, started to search the ground and clear a lane through the field. The rest of the advance guard deployed and increased the covering fire until page 94 the engineers signalled that the lane was cut, when the whole force charged through, the tanks leading. When there were infantry defences behind the minefield, the engineers would often ride the tanks to deal with any subsidiary fields. Although this method was likely to cost several vehicles, its value lay in the fact that the minefield would be breached before the defence could concentrate at the point of attack.

the panzer divisions break through the minefields, 31 august 1942—a trace from a 21 panzer division sketch map

The widest part of the British minefields, between the New Zealand Box and the south of Deir el Munassib where there were three main fields and one partially dummy, was encountered by 90 Light Division and the left flank of the Italian armour. The two panzer divisions broke through at points between Munassib and Himeimat where the fields merged into two, forming a belt some 200 to 1000 yards deep of scattered mines backed by a closely spaced and continuous line seven mines wide, then a clear lane of some 100 to 200 yards and a final continuous belt of three lines of mines. The second field, not so formidable, lay up to a mile and a half to the east. The route taken by 15 Panzer Division led it through two more isolated minefields, but 21 Panzer Division passed round the northern extremities of these fields, which were not completed.

The Panzer Army's advance fell behind schedule right from the start. Difficult going in the early part of the journey in the moonless period split Africa Corps, 15 Panzer Division diverging to the page 95 south and 21 Division to the north. This caused 15 Division to cut across the Reconnaissance Group's line of march, and at one time 33 Reconnaissance Unit found itself on the panzer division's left flank instead of on the right. Appeals by wireless to the panzer division's headquarters to check direction were useless as the division's wireless broke down until well after dawn.

On Africa Corps' left, 21 Panzer Division's infantry, in the lead, kept going on too northerly a course and met the minefields where they were thickest, while the tanks led by divisional headquarters swung back on the correct south-east line and hurried ahead to catch up with the sister division. The commander, Major-General Georg von Bismarck, a descendant of the Iron Chancellor, then returned to find the infantry, only to meet his death by British fire. The position of his grave and other evidence points to the cause of his death as mortar fire from 25 Battalion's patrol on Point 104.

The Italian armour, trailing Africa Corps' left flank, also travelled too far north and, halted by the minefields and British fire, found itself being crowded from the rear by 90 Light Division. By turning south along the minefields instead of cutting straight across, the Italians lost so many of their engineers that they had to call on 21 Panzer Division for assistance.

So great was the confusion, aggravated by the ground and air attack, that even Rommel himself was unaware of how his striking force was faring until dawn began to break and his liaison officers could bring him reports based on visual observation. He only knew that, at 5 a.m., when his troops should have been lined up south of Alam el Halfa ready to start the main assault on Eighth Army's rear, the formations were spread out over some miles of desert and well short of the first objective. The vanguards of the two German divisions were in fact by this time clear of the main minefields, but each gap was creating a bottleneck behind which columns were bunching, with supply vehicles strung out all the way back to the start line. A great deal of petrol had been used, many vehicles had either broken down or become bogged in soft sand, and others had been damaged by shellfire or bombing. The 21st Panzer Division alone had received the attentions of eighteen flights of bombers during the night. The only part of the mobile striking force to reach its dawn objective was the group from Folgore Division, which followed the wake of the Reconnaissance Group to occupy Qaret el Himeimat without much trouble.

On the British side, 7 Armoured Division's columns maintained field, anti-tank, and small-arms fire for as long as they thought wise, claiming—over-enthusiastically—as many as twenty-four page 96 enemy tanks knocked out. As the panzer engineers cleared each gap and let the tanks through, the columns fell back, in several instances breaking contact and withdrawing at speed for fear of being encircled. The German records include only one account of British troops being caught this night, when some tanks of 10 Hussars were destroyed and three men taken prisoner. Ground contact with the Panzer Army was uncertain at dawn, with conflicting messages coming from the various columns, but the Royal Air Force provided the army with reliable reports of the main enemy concentrations as an almost constant stream of reconnaissance aircraft and bombers flew over the southern front. A great degree of flexibility was shown this day when a heavy dust-storm arose in mid-morning, rendering many of the inland bomber airfields unusable; the air effort was then immediately switched to fighter and fighter-bomber sorties from the coastal landing grounds which were less affected by the dust.

By nine o'clock on the 31st, after regaining communication with most of the elements of his army, Rommel found that the Reconnaissance Group and 15 Panzer Division were some way to the west of Samaket Gaballa in light contact with British columns, 21 Panzer Division was still further west with its tail only just emerging from the minefields, while Littorio Division, the only Italian division to keep up with Africa Corps, was still threading its way through the last field on 21 Division's left rear.

The strength of the German enthusiasm for this offensive may be gauged from entries in the diaries of both the Panzer Army and Africa Corps. Whoever made the entries clearly had a feeling that the success of the operation was doubtful and suggested various alternatives, down to a complete and immediate withdrawal to the start line. Rommel, however, was not easily deterred, especially as his army had advanced so far without any real opposition from the British. He accordingly gave orders that the advance was to be resumed direct for Alam el Halfa, with the old telephone line that ran north-east as the centre line of the march. The immediate objective was to be the feature itself, to be assaulted by the two panzer divisions, with the Reconnaissance Group covering the eastern and south-eastern flanks and the Italian armour the western.

German reconnaissance aircraft had already reported this morning that an extensive infantry position existed on the Alam el Halfa ridge, but they appear to have failed to notice the hull-down tanks of 22 Armoured Brigade. No mass movement of tanks or transport vehicles had been observed in the British lines and this tied in with information Rommel had received from agents in Egypt, that the British planned to meet him with shellfire rather than an armoured counter-attack.

alam halfa, 31 august–1 september 1942

The extent of the disorganisation caused by the night march became apparent as soon as Rommel gave his orders. All units had taken the opportunity of the morning halt to refuel and start running repairs to their vehicles, and this delayed the assembly for the correct order of march. Then, about 11.30 a.m. the dust-storm thickened so that visibility at times was only a few hundred yards. Rommel's orders, given at 9 a.m., had allowed four hours for the panzer divisions to get ready, an exceptional length of time for German troops, but by 2 p.m. only 15 Division was on its way. On the left flank, Littorio Division, trying to make up for its previous tardiness, had pushed up into the area chosen by 21 Panzer Division for its assembly, thus bringing another hour of confusion as in the murk of the dust-storm Germans and Italians were sorted out, a task made no easier by the constant bombing and machine-gunning from the air. The management of the assembly was also delayed by the need for several changes in command caused by the wounding of Nehring and the death of von page 98 Bismarck; with 15 Panzer Division's commander taking over the Corps, both panzer divisions started the day with new commanders and a subsequent reallocation of the subordinate posts.

iv

As the Panzer Army was reorganising in the morning of the 31st, two of the light squadrons of 22 Armoured Brigade were sent out some three to four miles south of the brigade's position to observe and to act as a decoy to draw the enemy towards the prepared defences along Alam el Halfa. From ground and air observation and from wireless intercept, the Eighth Army had already begun to draw a fairly exact picture of the dispositions and composition of Rommel's forces, though for a time 21 Panzer Division could not be satisfactorily accounted for and the New Zealanders particularly feared that it was waiting in El Mreir ready to break through along Ruweisat Ridge.

Though some of 7 Armoured Division's columns were in contact with the Reconnaissance Group and appeared to have directed some harassing fire on the panzer formations, there was little action during the day until the dust-storm began to subside in the late afternoon. By this time 15 Panzer Division had progressed some miles along the eastern side of the telephone line and, as visibility increased, the light squadrons of 22 Armoured Brigade found themselves within range of the leading German armour and boldly took it on. The two-pounders of their Stuarts and Crusaders were no match for the enemy tanks and, after losing four tanks and a number of men, the light squadrons disengaged to drive rapidly back to the armoured brigade's protection.

Their role as a decoy was however not fulfilled, for 15 Panzer Division did not follow but swung out to the east. About the same time 21 Panzer Division, coming up on the west of the telephone line, appeared on the south of 22 Armoured Brigade, and it also began to turn east across the brigade's front. On this, the British commander ordered the tanks on his left flank to show themselves and open fire. The German tanks then turned straight at the armoured brigade's positions and joined in a tank versus tank battle that lasted until darkness fell. Twelve Grants, in spite of their hull-down positions against the Germans in the open, were knocked out, and the infantry protecting the tanks were caught in the cross-fire, losing several anti-tank guns and thirty-two men, of whom 21 Division claimed twenty as prisoners. The precise German losses in this action cannot be assessed but they were undoubtedly lighter.

As this engagement was in progress, 15 Panzer Division had turned to a northerly course and its leading tanks were probing at the infantry defences on Alam el Halfa. The headquarters of 10 Armoured Division, situated behind the minefield connecting company positions of 5 Royal Sussex of 133 Brigade, saw the enemy approaching and moved back some miles, but the German tanks turned back to the west along the front of the infantry defences and came in on 22 Armoured Brigade's left flank. The light squadrons guarding this flank fell back, exposing the brigade's gun lines and headquarters. With all the heavy tanks engaging 21 Panzer Division on the south, the brigade called for help from 23 Armoured Brigade, which sent two squadrons of Valentines to assist. Before this assistance arrived, darkness had begun to fall and both 15 and 21 Panzer Divisions broke off the battle and withdrew. However, of the two dozen tanks of 15 Division on the east of the armoured brigade, six did not fall back with the others, possibly because of empty petrol tanks or minor damage, but remained overnight across the rear of 1 Royal Horse Artillery's gun lines and less than a mile from the headquarters of 22 Armoured Brigade. Apparently these six were either unobserved or not recognised as enemy for they were not fired on until they were moving off in the dawn light next morning, when they lost two of the number in the process.

v

While the opposing armoured forces were joining battle well to their rear, the New Zealanders in their box remained relatively undisturbed except for the medium, field and anti-aircraft gunners. Save for a short period in the worst of the dust-storm, the noise of aircraft was almost continuous throughout the day as the Air Force attacked the enemy on the south and the Luftwaffe passed directly overhead on call from the panzer divisions to the east. The Bofors gunners of 14 New Zealand Light Anti-Aircraft Regiment, spread among the gun lines and headquarters areas, had one of their busiest and most successful days on record, with a final bag of three Stukas, and three Me109s shot down for certain and one ‘probable’. Two of the German pilots, parachuting in or close to the lines, were taken prisoner.

The field and medium guns were also kept busy firing on targets passed to them by patrols and OPs in the south, who watched the enemy columns moving through the line of the three depressions, the deirs Alinda, Munassib and Muhafid. Several of the observers were forced to move in closer to the Division's defences by probing page 100 patrols of the enemy, and at one period in the late afternoon there was a scare that an enemy force was advancing from the south against Alam Nayil. This alarm, and an earlier SOS from 5 Indian Division to fire on the west end of Ruweisat, brought all four field regiments and the medium guns into action with heavy concentrations on the threatened areas. There was in fact hardly a minute of the day when the New Zealand guns were silent, the artillery records showing that at least 12,000 shells—and probably considerably more—were fired between the time twelvebore was signalled and the fall of darkness on 31 August.

To the infantrymen in the box the day was comparatively restful. After their busy night, either in exchanging positions or taking part in the raids and patrols, followed by a long stand-to from the receipt of the alarm until well after dawn, the men were glad to snatch what rest they could in the heat, the dust, and the continual noise of gunfire and bombs.

Towards evening the Divisional Cavalry patrols and OPs, watching from the northern rim of Deir el Muhafid a long column of tanks passing to the south of the depression, had to fall back towards the south-east corner of the box as enemy infantry commenced to spread over the floor of the depression. The tanks were probably those of the Italian armoured corps, trying to catch up with 21 Panzer Division, and the infantry were part of 90 Light Division.

This movement gave General Freyberg considerable concern. During the day the columns of 7 Motor Brigade had fallen back to the east faster and farther than the plans anticipated and, although the Corps Commander had ordered them to return to give some cover to the south-east of the New Zealand Box, they had been prevented from doing so by the advance of Africa Corps. Their absence left the eastern front, occupied by 22 Battalion, and the south-eastern corner, where 23 Battalion had just settled in, uncovered except by the Cavalry patrols. This was the weakest part of the box, with no minefields extending beyond the perimeter belt.

On Freyberg's representations, Horrocks obtained permission from Eighth Army to bring 23 Armoured Brigade under 10 Armoured Division's command and ordered it to move south to cover the east of the box. By mid-afternoon the leading squadrons of Valentines were in the four-mile gap between 22 Armoured Brigade and the New Zealand Box. Later, two squadrons were sent to reinforce 22 Armoured Brigade.

With this arrangement Freyberg had to be content, but it was natural that he should remain anxious as, with the known reluctance of British tanks to move at night, an advance during darkness by the enemy armour followed by the customary dawn assault, with the rising sun behind it, would have been extremely difficult to resist. He issued orders that patrols were to maintain close touch with the screen of Valentines during the night and that the Division was to prepare for a night or dawn attack on the south and eastern defences of the box.

vi

August passed into September with a night of little incident. On the coast the Australians were busy with preparations for a daylight raid and neither invited nor received much attention from their opponents. The South Africans scored an unexpected success when a patrol of the Capetown Highlanders was ambushed by the enemy but fought its way clear with fourteen German prisoners in hand. On the rest of the western front, from Ruweisat to Alam Nayil, occasional gunfire and the movement of patrols were the only activity.

The New Zealand Division's southern front was more lively. A request had been made by the Corps Commander for ‘substantial raids’ to the south-west and south to harass the enemy's line of communication and to deter him from pressing overnight closer to the defences. This demand, however, reached the two New Zealand brigadiers rather late for the detailed planning they had come to see was the best insurance for success in such raiding, but on Freyberg's suggestion, they agreed to experiment with infantry parties working under close support from some of the Valentine tanks attached to the Division.

Shortly after dusk a strong infantry patrol set off from 18 Battalion's sector with three tanks of A Squadron, 46 Royal Tank Regiment, to raid Deir el Angar. On approaching the depression the Valentines went ahead at such a speed that the infantry were left behind and, encountering an enemy position, the tanks proceeded to drive through and over the trenches, firing all weapons and lobbing hand grenades until any immediate opposition had been silenced. The rest of the enemy in the vicinity, however, came to the alert and laid down an intensive mortar barrage between the tanks and 18 Battalion's party under which the infantry were forced to ground. The Valentines, with ammunition almost exhausted, withdrew through the mortar fire to rejoin the infantry page 102 and the whole party then returned to the box. No prisoners were taken, but identification gathered from the dead showed the sector to be occupied by a battalion of the Folgore Parachute Division.

Similar raiding parties with Valentine tanks were sent out to the south by 21 and 23 Battalions but both failed to find the enemy. The 23 Battalion patrol incurred casualties to men and vehicles in a minefield which had either been laid by the enemy or, more probably, by one of 7 Armoured Division's columns and not recorded.¹

Not very long after these patrols had returned to the box, a column of some fourteen vehicles, moving up between the routes taken by the 21 and 23 Battalion patrols, stopped within a few hundred yards of the defences held by D Company of 28 Battalion. The Maoris watched while the occupants, estimated at about 100 men, debussed to search for and lift mines. Once the troops were identified as hostile, the Maoris opened up with rifles and Bren guns, and sent up the light signal for the Vickers and 25-pounders to lay defensive fire on their sector. The enemy troops replied with automatics and anti-tank gun fire until the field guns got the range, when they withdrew, leaving behind the bodies of two men, three burnt-out trucks, and a light anti-tank gun. A patrol from D Company identified the dead as men from 90 Light Division and brought in the gun to add it to the battalion's arsenal of assorted weapons. The fact that the enemy could approach so close to the Division's defences before being observed caused considerable concern and brought orders from General Freyberg for a thicker screen of listening posts to be set across the front.

Throughout this night single enemy bombers droned overhead, occasionally letting loose a stick of bombs without causing much damage, while away to the south-east of the box Air Force flares lit the sky for long periods at a time and the rumble of heavy bombing could be heard distinctly by the New Zealanders.