Already on 29 June Hitler was worried that Bock’s Army Group Centre, where the advance was especially spectacular, would overreach itself.96 On 4 July he claimed that he faced the most difficult decision of the campaign: whether to hold to the original ‘Barbarossa’ plan, amend it to provide for a deep thrust towards the Caucasus (in which Rundstedt would be assisted by some of Army Group Centre’s panzer forces), or retain the panzer concentration in the centre and push forward to Moscow.97 The decision he reached by 8 July was the one wanted by Halder: to press forward the offensive of Army Group Centre with the aim of destroying the mass of the enemy forces west of Moscow.98 The amended strategy now discarded Army Group Centre’s turn towards Leningrad, built into the original ‘Barbarossa’ plan.99 The ‘ideal solution’, Hitler accepted, would be to leave Leeb’s Army Group North to attain its objectives by its own means.100 However, Hitler was even now by no means reconciled to the priority of capturing Moscow — in his eyes, as he said, ‘merely a geographical idea’.101
The conflict with Army High Command, supported by Army Group Centre, about concentration on the taking of Moscow as the objective, continued over the next weeks. Hitler pressed, in revised operational form, for priority to be given to the capture of Leningrad, and now included in the south the drive to the industrial area of Kharkhov and into the Caucasus, to be reached before the onset of winter. At the same time, his ‘Supplement to Directive No.33’, dated 23 July, indicated that Army Group Centre would destroy the enemy between Smolensk and Moscow by its infantry divisions alone, and would then ‘take Moscow into occupation’.102
By late July Halder had changed his tune about the certainty and speed of victory. Early in the month he had told Hitler that only forty-six of the known 164 Soviet divisions were still capable of combat. This had been in all probability an overestimation of the extent of destruction; it was certainly a rash underestimation of the enemy’s ability to replenish its forces. On 23 July he revised the figure to a total of ninety-three divisions. The enemy had been ‘decisively weakened’, but by no means ‘finally smashed’, he concluded. As a consequence, since the Soviet reserves of manpower were now seen to be inexhaustible, Halder argued even more forcefully that the aim of further operations had to be the destruction of the areas of armaments production around Moscow.103
As the strength of Soviet defences was being revised, the toll on the German army and Luftwaffe also had to be taken into account. Air-crews were showing signs of exhaustion; their planes could not be maintained fast enough. By the end of July only 1,045 aircraft were serviceable. Air-raids on Moscow demanded by Hitler were of little effect because so few planes were available. Most of the seventy-five raids on the Soviet capital carried out over the next months were undertaken by small numbers of bombers, scarcely able to make a pinprick in Soviet armaments production.104 The infantry were even more in need of rest. They had been marching, and engaged in fierce fighting, for over a month without a break. The original operational plan had foreseen a break for recuperation after twenty days. But the troops had received no rest by the fortieth day, and the first phase of the campaign was not over.105 By this time, casualties (wounded, missing, and dead) had reached 213,301 officers and men.106 Moreover, despite miracles worked by Quartermaster-General Eduard Wagner’s organization, transport problems on roads often unfit even in midsummer for mechanized transport brought immeasurable problems of maintaining supply-lines of fuel, equipment, and provisions to the rapidly advancing army. Supplies for Army Group Centre required twenty-five goods trains a day. But despite working round the clock to convert the railway lines to a German gauge, only eight to fifteen trains a day were reaching the front line in late July and early August.107
