Page 7

Chapter 2: Lull in the Desert War


EVEN the experts were baffled in mid-1941 by the strategic problems of the British Commonwealth in its lonely struggle with the Axis powers, perhaps soon to be joined by Japan with baleful repercussions. Planning to win the war was at this stage like planning to win a sweepstake. All that could be done was to buy more and more tickets and risk going bankrupt. Some tickets were on blockade, some on sabotage and rebellion in countries in enemy hands, some on a great bomber offensive, and some on the entry of the United States on the British side. None was on the entry of Russia, which was not thought likely to bring any but temporary advantages.1

The security of the bases in the United Kingdom and Singapore by the accepted strategy was prior to that of the Middle East, but it was mortgaged, as General Dill thought, by Mr Churchill to finance dubious Middle Eastern enterprises. In this connection, though Wavell’s intimates recognised that he was ‘always ready to take a chance’,2 Churchill gained the opposite impression and it was a grave blow to Wavell to find that in one case – the Iraq revolt – his own judgment was wrong.3 But time adds perspective to these clashes of opinion and personality. The British position in the Middle East was saved, as it happened, not by Churchill or Wavell, but by Hitler. Overwhelming German intervention in the Middle East was possible and indeed warmly recommended by Hitler’s naval chief, Grand-Admiral Raeder, and by Reichsmarschall Goering himself. But Hitler instructed the Wehrmacht instead to conquer Russia.

This was as Churchill suspected, but his hopes needed no such rich nourishment and he refused to wait. He had ventured to

Page 8

challenge the German colossus on the mainland of Europe with no allies but Greece and a Yugoslav faction. Scarcely had the last British troops left Greece, however, when a new desert offensive was urged on Wavell, concurrently with an invasion of Syria and, if it lasted long enough, with the battle of Crete. Forces and supplies were shipped round the Cape early in May and tanks and aircraft rushed through the Mediterranean at Churchill’s instigation and at immense risk, with injunctions to Wavell to put them to good use at the earliest moment. The risks taken made even the appearance of delay smack of base ingratitude. ‘All our hearts at home’, Churchill says, were ‘set on beating Rommel in the Western Desert.’4 Great pressure (‘undue’ according to Dill) was exerted on Wavell and he agreed to attack ‘before he was fully prepared’.5 Five extra days were grudgingly conceded by Whitehall for crews to get used to their new tanks and the offensive, code-named BATTLEAXE, opened on 15 June, eight days after the invasion of Syria began.

Despite heavy commitments elsewhere, Wavell had been anything but hesitant in his handling of the desert operations. An Axis assault on Tobruk had come to a painful halt early in May. Then, in a brief effort (BREVITY), he tried in the middle of the month to regain the frontier area and perhaps relieve Tobruk. All that could be seized, however, was Halfaya Pass and even this vantage point was lost at the month’s end to a German counter-thrust. BATTLEAXE was intended not merely to recover this lost ground but, as Churchill wrote on 27 May, to ‘inflict a crushing defeat upon the Germans in Cyrenaica’.

Such a victory over German forces was a consummation greatly to be desired; but its ingredients were essentially military, and of these Wavell was the better judge. He had hoped to mount BATTLEAXE before the full weight of the newly-arrived 15 Panzer Division could be brought to bear, but this hope faded early in June. He voiced misgivings, too, about British equipment in the light of current reassessments: the cruiser tanks were unreliable, the infantry tanks too slow, the armoured cars too lightly armed and armoured, and the enemy anti-tank guns unexpectedly powerful. It was therefore a gamble to attack with fewer tanks and perhaps fewer infantry than the enemy; but Churchill refused to see BATTLEAXE in this light and limit his hopes of success.

Page 9

BATTLEAXE Operation, 
15–17 June 1941

BATTLEAXE Operation, 15–17 June 1941

The desert operation fell far short, in the event, of Churchill’s expectations and short even of the limited success for which Wavell hoped. The infantry tanks (‘Matildas’) of 4 Armoured Brigade destroyed or damaged 50 out of 80 of the tanks of 15 Panzer Division in one morning and overran many guns; but 7 Armoured Brigade (with cruiser tanks) was much weakened in an outflanking move and failed to hold a strong counter-thrust. On the third day Wavell intervened in person and allowed the commander of the Western Desert Force, Lieutenant-General Beresford-Peirse, to call off the operation. The force withdrew to Sidi Barrani in ignorance of the damage inflicted on the enemy, which was not nearly so heavily outweighed by the British losses as was thought at the time. Freyberg wrote on 9 October: ‘We suffered decided reverse and lost large number of tanks. After battle several comds were sent to other jobs.’ But dismissals and courts-martial ensued on the German side and Rommel was anything but pleased with several of his senior officers.

To Churchill, however, it was a ‘most bitter blow’ and he received it alone at Chartwell, where he ‘wandered about the valley disconsolately for hours’.6 Wavell for his part summed up the situation quickly and accurately and reported next day to London that ‘no offensive in the Western Desert would be possible for at least three months’, thereby rubbing salt into the wound of

Page 10

Churchill’s disappointment. A few days later Wavell agreed with his naval and air colleagues to accept the commitment of holding Tobruk at least until the autumn, though the garrison could not now hope for early relief.

Wavell was now tired out and in need of a rest. The bombardment by memoranda from London which Admiral Cunningham (also a target) has described as ‘singularly unhelpful and irritating in times of stress’,7 on top of the inescapable cares of Wavell’s vast command, was a serious distraction. Air Chief Marshal Longmore had been abruptly dismissed early in May and Wavell suspected that Churchill had for some time been itching to get rid of him too. The failure of BATTLEAXE settled the issue and the decision about the future of Tobruk was Wavell’s last important act as the Middle East Commander-in-Chief.

At the end of June he changed places with General Sir Claude Auchinleck, Commander-in-Chief India, and in a moving message of farewell on 4 July he expressed his thanks to the troops who had served him ‘so well and loyally’. New Zealanders not long back from Crete read of the failures and setbacks when ‘you have been outmatched in numbers and equipment, never in fighting qualities or endurance’. For those serving in rear areas there was a special word of praise for work ‘essential to success in battle’. To this General Freyberg replied ‘on behalf of all ranks of the NZEF’ expressing the ‘genuine sense of personal loss’ felt by the New Zealanders, which Wavell gracefully acknowledged on 7 July.


That Churchill, most articulate of men, should prefer the fluent Auchinleck to the taciturn Wavell is not surprising. Yet the choice was in one way a curious one; for Auchinleck had already incurred Churchill’s displeasure by seeming to ‘play too much for safety and certainty’ in the Norwegian campaign,8 a fact of which Churchill was soon to be reminded. General Dill was alarmed not so much by the change as by its implications and hastened to put his views to Auchinleck in a letter of 26 June. In this remarkable document Dill absolved Wavell from blame for attacking in Syria with inadequate resources and for mounting BATTLEAXE with undue haste – ‘except in so far as he did not resist the pressure from Whitehall with sufficient vigour’. On the other hand he allowed that ‘pressure from those who alone see the picture as a whole and carry the main responsibility may be necessary’ and that political

Page 11

advantages sometimes outweighed military drawbacks in particular projects. Auchinleck would not have all the resources needed for his ‘great task’; these would come later. ‘But in the meantime we have a grim fight to fight and we cannot afford hazardous adventures’, Dill wrote. ‘So do not be afraid to state boldly the facts as you see them.’9 In the terms of Wavell’s dismissal, written on 21 June, Churchill wanted ‘a new eye and a new hand’ in the ‘most seriously threatened’ Middle Eastern theatre;10 but Dill was evidently anxious that these should not be guided by Churchill alone. Next morning the nature of the threat to the Middle East changed dramatically when BARBAROSSA (the invasion of Russia) opened, providing welcome though possibly short-lived easing of German pressure elsewhere.

Wavell’s summing up after BATTLEAXE was promptly forgotten. Auchinleck reached Cairo on 30 June and the very next day Churchill urged him to ‘renew the offensive in the Western Desert’, if possible before the fighting in Syria ended. For at least six weeks, and possibly three months, almost the full weight of Germany would be turned against Russia. What broad happy vistas this opened up: Cyrenaica, then Tripolitania, then Sicily or French North Africa! But the cupboard was bare. BATTLEAXE had expended the prime requisites for a major offensive, and the new commander-in-chief, with Dill’s backing, was not slow to point this out.

Auchinleck replied on 4 July in terms which contrasted sharply with Churchill’s urgency. British influence in Syria would have to be secured and in Iraq ‘re-established’, then came the defence of Cyprus – in fact all that was needed to provide a firm base. Only then could a desert offensive be ‘contemplated’, though he estimated that it would call for two, and perhaps three, armoured divisions. This did not sound at all like the kind of ‘new eye’ and ‘new hand’ that Churchill had intended and he set out his argument in some detail on 6 July for an offensive not later than mid-September and preferably much earlier. Prospects would be favourable by the end of the month, would not improve during August, and might diminish drastically in September – or sooner if Russian resistance collapsed. Churchill was anxious also about the ‘offensive value’ of the Tobruk garrison – a pet theme of his11 – in two months’ time. But Tobruk was no Gerona and the Australians found it extremely hard to make any deep impression in the enemy lines from May onwards.

On 15 July Auchinleck stated conditions for the success of a desert offensive which, if accepted, promised to delay the opening

Page 12

until the end of the year. Even Dill was somewhat taken aback by this and the London Chiefs of Staff tried various inducements. Could Auchinleck not start the operation at the end of September, they asked on 19 July, if they sent him another 150 cruiser tanks at once and another 40,000 men? Otherwise they could not afford this diversion of scarce shipping. With these tanks, Auchinleck answered, he might mount a limited offensive to relieve Tobruk in mid-November – a fortnight earlier if trained crews were also provided. If he could also get 150 more American tanks and certain heavy transport he might undertake a major offensive to drive the enemy out of North Africa. As things stood, he pointed out, he would not have even one armoured division ready for action by the end of September.

To the Defence Committee this seemed agonisingly slow. Now was the time to strike, or within the next few weeks, while the Germans could least afford diversions from their vast effort in Russia and while they were hard put to it to maintain their troops in the desert. There was only one way out of the impasse: would Auchinleck and Air Marshal Tedder, Longmore’s successor as AOC-in-C, come to London and talk it over? This question was put to them on 23 July and the two reached London on the 29th.

There and at Chequers the argument continued. There was indeed much to be said for an early offensive and Auchinleck heard it over and over again: the political and military need to help Russia, the unlikelihood of a land attack from the north (from Turkey or the Caucasus) before mid-September, the strain of supplying Tobruk and the value of a concurrent sortie by its garrison, Axis supply troubles in North Africa, and so on. But he remained unmoved and with impressive dignity and eloquence presented the case for a November opening for the coming offensive, now called CRUSADER. He outlined the immense labours which must precede it in office and workshop, on the parade ground, and in the desert itself. Assessing strength by counting tanks, guns and heads was of little use. Training at all levels was all-important; to skimp it was to invite disaster, as BATTLEAXE had shown.

It was Churchill who yielded in the end, though unconvinced, and Auchinleck and Tedder returned to Cairo with a promise from the Defence Committee that 22 Armoured Brigade (from 1 Armoured Division) would be sent to the Middle East as soon as possible. In view of this Auchinleck had agreed to start CRUSADER on 1 November.

There remained, however, a marked difference between Auchinleck’s and Churchill’s views of the offensive. Churchill expected it to open ‘a continuing path’ leading, as a matter of

Page 13

course, to Tripolitania and if possible to French North Africa or even Sicily. But Auchinleck was still inclined to see it in terms of security rather than gain. It was a tight situation which faced him. Malta was cut off after tiger12 from supplies from the east and the Tobruk supply line was expensive in small ships and air cover. The Fleet bases at Alexandria and Haifa, the Suez Canal, and Suez itself (the chief point of entry for Middle East supplies) were subject to air attack and a concentration of bombing against any one of them – particularly Suez – was much to be feared. Then there was the danger that the Germans might come down through Turkey or the Caucasus.

Auchinleck’s sentiments were shared by Admiral Cunningham and Air Marshal Tedder, to whom CRUSADER’s greatest attraction was the expectation that it would yield airfields in Cyrenaica and thereby greatly ease the strain on Malta. The importance of this tiny and much-bombed island in affording security to the base in Egypt was out of all proportion to its size. Only attacks from Malta on Italian shipping would allow the British forces with their 13,000-mile supply lines to build up faster than the Axis forces in North Africa. To sustain these attacks demanded a great supply effort, with massive naval and air support. Air raids on Malta slackened during May and even more in June and July, because of BARBAROSSA and as a kind of rebate from the loss of Crete, to which some Luftwaffe units were transferred from Sicily. But success against enemy shipping could best be achieved by surface striking forces based on Malta and these, after Crete, could not be afforded. In June and July sinkings therefore decreased and more Axis supplies got through to Libya. Not for a moment could the Middle East Commanders-in-Chief afford to lose grip of this critical situation.


In the rarefied atmosphere of high strategy the New Zealand Division, while recuperating from Greece and Crete, was mentioned, in a manner unacceptable to Freyberg, only as a possible relief for the garrison of Tobruk. This fortress was, like Malta, a symbol of gallant defiance maintained only by straining scarce shipping resources almost to breaking point. As a festering sore in the side of the German-Italian army in Libya it was well worth the effort; but Churchill demanded even more. He had been upset that the BATTLEAXE plan did not include as ‘an indispensable preliminary and concomitant’ a sortie from Tobruk,13 perhaps not knowing of

Page 14

the hard fight of one brigade of the garrison in support of BREVITY, nor of the continual exertions by the Australians to drive in the dangerous enemy salient at Ras el-Medauuar, nor yet of the trouble the commander (Major-General Morshead) had had to prepare another brigade to link up with the Western Desert Force in the second stage (which failed to eventuate) of BATTLEAXE. The garrison was not nearly strong enough to break out on its own and a premature move would have been dangerous if not disastrous, as is only too clear from subsequent fighting. By the end of June only 1000 yards of the salient had been regained on a front of a mile and a half and Morshead had to keep three brigades at full stretch to man the 30-odd miles of perimeter, each being relieved by the fourth from time to time to find what rest it could, though there was no haven beyond reach of enemy guns and dive-bombers.

When BREVITY and then BATTLEAXE failed, the relief of Tobruk became a dream of the future. Most of the garrison had been committed to action on 8 March and besieged since 10 April and ‘a wave of pessimism swept over the defenders’,14 though with digging, wiring and mining and almost incessant patrolling they soon regained their confidence. By night no-man’s land was theirs; but they could not fail to note the increasing activity of enemy guns and aircraft, while their own guns for lack of ammunition had to control their tempers.15 This and the heat, the brackish water, and the unvaried diet caused cumulative strain and sickness, often of a kind not adequately reflected in medical reports. As the post-BATTLEAXE lull continued into July and operations in Syria ended, Lieutenant-General Blamey, now Deputy Commander-in-Chief, Middle East Forces, proposed that the garrison should be relieved by sea, by the New Zealand Division or 1 South African Division.

Middle East was not so amply supplied with troops, however, that a whole division could be replaced without widespread repercussions. The relief of 18 Australian Brigade (needed by its parent division16) by 1 Polish Carpathian Brigade was already arranged; but who was to take over from Morshead’s division? The question was settled in Auchinleck’s absence17 by a Cairo conference of 2 August at which 70 Division, formerly 6 (British) Division, was chosen.

Undoubtedly connected with this decision was a sudden, brief flurry of activity among the New Zealand authorities at Maadi and Helwan on 4 August. A warning order was hurriedly issued for

Page 15

Divisional Headquarters and 4 Brigade to move to Syria at short and rather mysterious notice. Advanced parties would report at Cairo railway station the same evening. ‘Destination and action at destination will be issued later’, the order cryptically added. By the afternoon it was cancelled and the fuss ended. The Poles duly relieved the Australian brigade and the Indian cavalry regiment in Tobruk in mid-August; 70 Division prepared to relieve Morshead’s 9 Australian Division.

Opponents of the relief fought a determined rearguard action, but Blamey and three successive Australian governments insisted and all but a battalion and a half of the Australians were brought out by the end of October. British sources point to Australian domestic politics as the villain of the piece and Churchill is particularly outspoken. Yet it is hard to see how a major role for the garrison could have been allotted troops who had been in action for eight months and besieged for seven of them. Even the last stage was vigorously opposed, though there was little to be said for embarking on CRUSADER with a garrison of bits and pieces – British, Poles and disappointed Australians.18


Crusader was to take place in a corner of the world’s greatest desert, the desert which stretches from the Nile for 3000 lonely miles westwards to the Atlantic and for 1500 equally desolate miles southwards nearly to the Equator. In current Army terms there was first the Western Desert as far as the Egyptian frontier, then the Libyan Desert, with the great Sahara to the west and south of it, though in truth from Nubia to Mauretania was all one wasteland, hostile to life and miserly with its treasures.19 These could not sustain modern armies and everything was imported: the men, their machines and supplies, and the issues they contested – all, that is, but water, found at scattered points along the coast and at rare inland oases.

Among the few fertile parts are the Green Mountain (Jebel Akhdar) of Cyrenaica, bringing rain and vegetation in season to 200 miles of coast from Benghazi to Bomba. There were found – or so said the Romans – the garden of the Hesperides and the dark waters of Lethe. But this ancient granary with its golden memories is small indeed alongside the endless miles of semi-desert which stretch across the base of the Cyrenaican hump, merging into deep sand dunes which seal off the southern boundary 100 to 180 miles

Page 16

from the coast. This barren and almost impassable dune-land, the Libyan Sand Sea, is flanked on the north and north-east by the two large oases of Jarabub (in Libya) and Siwa (inside Egypt), 150 miles inland, and east of these lie the salt marshes of the Qattara Depression, narrowing the desert eastwards until at El Alamein it is no more than 30 miles wide. The Western Desert is thus a funnel starting 60 miles from Alexandria and opening westwards into the steppes of Cyrenaica, 300 miles long by 150 wide. In this tufted semi-desert with limestone outcrops, a few twisting wadis dry for all but a few days each year, and here and there an Arab tomb or other vague landmark, the desert armies were destined to meet.

Between El Alamein and El Agheila they would have to share a 600-mile open flank in this semi-desert with seldom a serious restraint of climate or terrain on the movement of mechanised forces – a tactician’s paradise indeed, but only for those with strong nerves. ‘Going’ in the Sollum-Tobruk area, for example, varied from impassable escarpments (mostly parallel to the coast at a distance of a few miles) to firm, smooth surface well inland which allowed mile after mile of comfortable top-gear driving. The coast strips below and just above the escarpments were inclined to be treacherous, with soft sand and broken ground entailing much gear-changing and many defiles. Here, too, the winter rains carved their wadis up to 30 miles inland and were survived for days or weeks by patches of mud or swampy ground (and hints of greenery), compounding the deception of soft sand and calling manpower all too often to the aid of mechanical horse-power. Sandstorms could at other times halt all movement in a wilderness of discomfort, though the wind more often achieved no more than minor irritation from sand scudding along the surface, clogging food, hindering the maintenance of machinery (especially aircraft engines), and fraying tempers.

The main escarpment on the Egyptian side took shape south of Sidi Barrani and curled and rose for 50 miles towards the frontier until at Upper Sollum it became a 500-foot cliff overlooking the sea, mounted only by a tarmac road, Sollum Pass, or by a lesser route, Halfaya Pass, four or five miles to the south-east. West of Bardia this escarpment broke up into smaller ones which in turn merged into two south of Tobruk. In the autumn of 1941 what these lost in height they gained in tactical stature as they approached that fortress. To the south terrain features were so modest and rare that the frontier mostly followed the 25th meridian. The Italians had put up a 169-mile barbed-wire fence embedded in concrete on the

Page 17

Libyan side of the frontier to restrict the wanderings of the Senussi Bedouin, but by mid-1941 this had become dilapidated and of little significance.


In the summer and autumn of 1941 the frontier divided the opposing forces except for the British garrisons of Tobruk and Jarabub and patrols of the Long Range Desert Group which ranged deep into Libya, offset by a small but important enemy encroachment on Egyptian territory at Halfaya. The desert armies clung in the main to the coast, served by the coast road and on the British side by the railway to Mersa Matruh. On the enemy side a 47-mile road by-passing Tobruk (Strada del’Asse or Achsenstrasse) was laboriously constructed by 6000 Italians and opened on 9 August, saving enemy vehicles a rough cross-country journey. Only reconnaissance, training manoeuvres, or major operations would tempt either side for long into the desert hinterland. Though the desert was one vast highway for motor transport it charged a high price in lorries, fuel and supplies, all of which were husbanded for operational demands.

Even near the coast the desert miles drank petrol copiously and posed many awkward problems for an army like the British in which everything on the ‘Q side’ stemmed from Railhead. By this European doctrine Railhead was a benevolent institution established as near to the scene of fighting as comfort and convenience allowed. From Railhead came the troops, tanks, guns and other vehicles of the fighting formations. From it also came the supply lorries, like dusty or muddy pearls strung together in road convoys bearing food, ammunition, fuel, engineering equipment, medical supplies, comforts – whatever was needed. Back along the network of roads came the ‘empties’, the wounded, and, if Fortune smiled, the prisoners.

Thus it was in theory; but the desert decreed otherwise. Railhead there was indeed, by a happy conjunction of foresight and luck; but it was 140 miles by road from Matruh to Sollum, 190 rough miles to Jarabub, some 300 by a roundabout route to Tobruk, and more than 600 to the far corners of Cyrenaica. The conquest of this vast area by manoeuvre on the open flank would inevitably take the British forces far beyond reach of the usual supply services – the two echelons of RASC transport which normally travelled between Railhead and the fighting units.

Each mile the railway advanced westwards from Matruh would save many dozens of supply lorries and the extension of it towards the frontier became a matter of urgency. A tricky eight miles up an escarpment south of Matruh and on to the plateau had been

Page 18

surveyed by 9 NZ Railway Survey Company and formed (but not completed) by New Zealand technicians and Arab labour by February.20 Work did not start again until 1 June, but this time the task was tackled in great earnest and gained increasing momentum. After BATTLEAXE indicated that none but an offensive on the largest scale offered hope of success, more and more labour was applied and by 15 September the line reached Mohalfa (formerly ‘Charing Cross’), gaining 17 valuable miles. The New Zealand Railway Construction and Maintenance Group with two construction companies (the 10th and, after 22 September, the whole of the 13th) and splendid Indian labour pushed the line westwards at a rate which by 20 October reached the astonishing average of two miles per day – bewildering to desert navigators who were apt to find their calculations wildly astray on their return journeys. The aim was to set up a new railhead and have it working as far west as possible before CRUSADER opened; but there were some who felt the New Zealanders had raised their sights too high. All criticism, however, was happily confounded when the new line reached Misheifa, 93 miles from Matruh, on 8 November and the new railhead with nine miles of sidings was opened for business on the 15th. The saving in lorries thereby has been estimated21 at 4000–5000 and the railway cut the task of maintaining the large forces earmarked for CRUSADER, at least in the early stages, down to manageable proportions.

In one way, however, the railway robbed Peter to pay Paul, since the locomotives needed more water than the whole desert army would drink, and demanded a huge increase in the supply. In fifty-six days seven pumping stations and ten large reservoirs were built and 145 miles of water pipeline laid – a vast programme completed, like the railway, in the nick of time to meet the needs of CRUSADER. By 13 November water was reaching the railhead in adequate volume, much of it piped 270 miles from Alexandria.22

These great construction works went hand in hand with other capital outlay on the forthcoming offensive: airfields in the Canal Zone for heavy bombers, forward landing grounds for light bombers and fighters, and roading and similar works as far afield as Syria. Thus the effort of building up and maintaining a force of little more than 100,000 in the forward area engaged many times that

Page 19

The Battle Area

The Battle Area

number throughout the Middle East, to the astonishment and dismay of London authorities who found it hard to grasp the administrative limitations imposed on operational planning in an undeveloped land.


On the frontier July and August passed quietly, though in the coast sector three columns (fait, hope and char in current Signals jargon) took turns at playing a dangerous game of provoking an enemy who in his strong arc of defences covering Halfaya and Sollum ‘held all the aces.23 Gunners would push forward a single 25-pounder to snipe by night and later, as the technique improved, by day, attracting as intended far more fire in return than they expended themselves. On the plateau above the escarpment light mobile forces watched and waited while the enemy toiled with pick and shovel, digging, concreting, wiring and laying mines to extend the defences in a series of battalion strongpoints 30 miles south-westwards from Sollum to Sidi Omar, a strong arm to ward off British intervention long enough to allow the Axis troops, when the stage was set, to capture Tobruk.

This was a project on which Rommel had set his heart, but the cold facts of administration were against it. He had no counterpart to the railway from Alexandria and no sizable port nearer the front than Benghazi. Transport aircraft did what they could to make up deficiencies but the supply situation was always uncertain and (largely because of His Majesty’s submarines) at times critical. So the Tobruk operation was several times postponed and the prospect of invading Egypt faded into the distant future.

General-of-Panzer-Troops Rommel was nominally under the command of the Italian General Bastico, Commander-in-Chief of the Italian Armed Forces in North Africa, but Bastico’s few attempts to assert his authority proved singularly unsuccessful. It may have been some consolation, however, to know that even the Army High Command in Berlin24 also failed. A large independent headquarters and liaison staff was sent to Libya under a General Staff officer, Major-General Gause, who had orders not to place himself under Rommel’s command but to report direct to Berlin. But Rommel’s ‘morbid ambition’ (as Gause described it) and Hitler’s continued support proved too much even for OKH. In mid-August Rommel absorbed Gause and his staff into a newly-formed headquarters, Panzer Group Africa, and Gause acted from then onwards as his Chief of Staff. This was virtually an army command, including as it did the German Africa Corps,25 21 Italian Corps, and 55 Italian

Page 20

Savona Division (in the frontier area); but its only mobile troops were 15 and 21 Panzer Divisions. The 20th Italian Mobile Corps, not yet battle-worthy, remained directly under Bastico.

The non-mobile troops, including a German division (Division z.b.V.26 Afrika) in course of being formed from independent regiments and battalions, had plenty to do either on the Tobruk front or the frontier line, and in August and September several small actions were fought to tighten the ring around Tobruk; but the panzer troops languished in the summer heat. Their whole background of doctrine and training had been directed towards mobile rather than static operations.

A limited operation of some sort was clearly called for and in mid-September it was provided: a raid, code-named SOMMERNACHTSTRAUM,27 by the bulk of 21 Panzer Division to overrun a supposed British dump some miles south of Sollum and strike heavy blows at two British groups thought to be defending it. As the time approached Africa Corps strongly suspected that the ‘dump’ was empty and doubted whether much damage to the British groups would result; but for Rommel the operation was essentially action as opposed to inaction and had irresistible appeal. In the pale light of the last-quarter moon early on 14 September, he went forward personally with one of two battle groups of the division which threaded their ways through the minefields while 3 Reconnaissance Unit swung widely southwards and then north-eastwards to simulate a third attacking force. Fighter aircraft brought forward for the occasion to Gambut airfield stood ready to help.

The British covering forces, however, were amply forewarned by the roar of engines and tank tracks in the still desert night. The leading battle group came on fast soon after dawn but the British withdrew even faster, though one South African armoured-car commander was able to report that he was ‘lying a close second to a German tank’ in an exciting race.28 By breakfast-time enemy tanks had reached Hamra, 20 miles south-east of the suspected dump, and the wild-goose chase reached Sofafi, another 20 miles eastwards, in the course of the afternoon. There the enemy halted, the tanks refuelled in dangerously close order, and the RAF and SAAF bombers caught them in the act. There had been no parallel movement in the coast sector below, where the sniping gun, well forward in its minor war of attrition, ‘broke all records by scoring 5 for 225 before the breakfast adjournment’ – five rounds fired for 225 returned.29 But after dark even this coastal force withdrew,

Page 21

leaving a small rearguard at Buq Buq and sending another nervously south-westwards from Sidi Barrani to meet the enemy armour. Behind them the water point at Buq Buq was needlessly demolished.

The night of 14–15 September had on the British side an exciting uncertainty and various detachments prepared themselves as sacrificial offerings if the Germans continued the advance. As far back as Mersa Matruh 5 South African Brigade sent a battalion forward as a delaying force. But no sacrifice was demanded and on the 15th the German armour headed westwards, pausing only to shake its fist at armoured cars which followed insolently close.

The dream had ended. How happy the awakening was can only be surmised from the German documents and contemporary comment was understandably guarded. There was some consolation from the capture of a South African office truck with three prisoners and some interesting papers (as against 12 Germans and 16 Italians captured). But SOMMERNACHTSTRAUM came at an unfortunate time from the viewpoint of German Intelligence. One of its main objects was to reassure Rommel that no British offensive was impending and thereby free him from worry about what the British might do when he attacked Tobruk. All Rommel could learn from the captured papers was that the British covering forces were slender and planned to move back quickly if seriously threatened. A week or so later he might have learned of a scheduled strengthening of these forces to protect forward depots and landing grounds for CRUSADER. But Rommel liked to ride his dreams bareback and was only too eager to accept whatever support SOMMERNACHTSTRAUM offered for the view that British moves would not clash with his attack on Tobruk.

As a fillip to German morale, which was also intended, the Dream can have paid but a small dividend, the British mobile troops being adept and pugnacious with their guns and too slippery for effective reprisal, while the ‘carpet bombing’ which the Germans experienced for the first time at Sofafi was anything but reassuring. It did much damage to the panzers and very nearly killed Rommel himself. The tank strength of 5 Panzer Regiment dropped drastically from 110 to 43, a difference of 67 which was made up so slowly that it was not until November, on the eve of CRUSADER, that the former total was reached. Some of the 67 may have been in workshops for routine overhaul but the majority were probably damaged in some way or other, though only two tanks were abandoned on the field.

In aircraft the situation was even worse. Though the enemy got slightly the better of the fighter clashes, the balance swung heavily against him after a raid on the crowded Gambut airfield. The Panzer Group war diary plaintively records on 15 September that

Page 22

Fliegerführer Afrika, the German Air Force commander, telephoned to say that no aircraft at all would be available until 1 p.m. next day; all fighters at Gambut had been put out of action and bombers had to have fighter support. As a crowning indignity a number of Stukas manned by Italians landed within the British lines and were captured intact.

When the fuss died down preparations for CRUSADER went ahead quietly. The 4th Indian Division under Major-General Messervy assumed command of all troops in the forward area and its 11 Infantry Brigade took over the coast sector later in the month. The 7th Armoured Division also moved up inland, strengthening the covering forces and promising heavy punishment to any similar reconnaissance-in-force the enemy cared to attempt. As a further insurance the New Zealand Division began to assemble 180 miles behind the front.