Danube and Sava, on the W. the line of the Drina, a river not
very broad, but rapid and difficult to bridge, these rivers form-
ing the actual frontier. 1 Initially then the assailant must begin
by forcing a river barrier, whatever the direction of his advance,
and his power to do so depended: (a) on the communications on
his own side of the river, which would enable him to collect the
troops and material for forcing the passage; (b) on the barrier
itself; and (c) on the defensive positions available for the Serbs
inside the barrier. Taking all these factors together, an attack
on the northern barrier offered the best chances of success. It
was the course chosen by Prince Eugene in 1717 and by Macken-
sen in 1916, and it was that which the Serbs themselves regarded
as most likely to be taken in August 1914. The reasons which led
the Austrian commander to choose differently will be considered
presently. More important than the choice of a point at which the
barriers could most easily be forced was the choice of a direction
for the subsequent advance. The objectives were plain enough,
the defeat of the Serbian field army and the occupation of the
most important part of the country. Without the attainment of
these objectives, Austria could not pretend to have conquered,
or even to have defeated, Serbia. Both, however, were attain-
able by an invasion of the Belgrade-Pozharevats-Kraguyevats-
Krushevats region, conventionally called the Morava valley.
This was the richest part of the country, and the centre of all
its communications. Here and here only could the Serbian
army be definitely brought to action under conditions in which
the Austrian superiority of numbers would be effective.
This region could be approached from the N. (Belgrade or Danube-Pozharevats front), from the N.W. (Sava and Lower Drina front), from the W. (Middle Drina front) or from the S.W. (Upper Drina front). 2 The first named was the most direct route. For the forcing of the Danube 3 five railheads on the Hungarian side were available, for the prosecution of the advance the Orient railway (Belgrade-Constantinople) and' various branches of it. The country was rich in local supplies, populous, and relatively well provided with roads, and the general trend of the natural features, governing that of the lines of communica- tion, was from N. to S. On the other hand, the defender would of course be best prepared for attack on this front, and the Danube obstacle was here most difficult to overcome, in spite of the command of the water ensured by the Austrian monitors. 4 Next in importance were the lines of advance from the Sava and Lower Drina, and especially those from the Sava, since the Drina valley was totally destitute of railway approaches on the Hungar- ian side. In the angle of the Sava and Drina lies the fertile plain of the Machva; many favourable points exist on the Sava for the forcing of a crossing, and the town of Shabats in particular offers a choice of tolerable roads leading S. to the slopes of the Tser (Cer) range S.W. to Valyevo, and W. to Arangyelovats, the last-named route conducting an invader to either Palanka or Kraguyevats in the Morava region. In conjunction with such an advance, an operation on the left, over the Lower Kolubara, sup- plied from the Sava, would lead to the rear of Belgrade, while on the other flank a successful operation by troops organized for mountain warfare by Valyevo over the Kolubara and Rudnik ranges to Gorni Milanovats and Chachak would threaten the rear of the whole Morava region. As a way of turning the great obstacle of the Danube front, the lines of advance grouped on the axis Shabats-Arangyelovats possessed tempting advantages. But, considered from the point of view of supply, the disadvan- tages were very serious. Not only was this the longest route
'In its upper course the Drina passes inside the Austrian frontier, and direct advance into Serbia is therefore possible from the region of Vishegrad. But this roadless, mountainous region is quite unfitted for the movements of large forces.
1 Attack from the stretch of the Danube front lying in the moun- tains between Weisskirchen and the Iron Gates was regarded by both sides as outside the limits of probability.
- The lowest reaches of the Sava are usually considered as belong-
ing to this front.
4 Six in number, armed with 4-y-in. guns, and armoured.
6 The small line connecting Shabats with Loznitsa was of no importance except in the case of stabilized trench warfare m the Machva.
between the railheads of deployment and the final objective, but it was destitute of railways 6 E. of the Kolubara. West of that river, indeed, the line Valyevo-Arangyelovats-Palanka lay in the very axis of the march, and a branch-line connected this with Obrenovats on the Sava, where water-borne supplies could be unloaded. But the barrier of the Kolubara and the ridge behind it had to be mastered before this branch could be used, and it was certain that the Serbs could assemble in time to fight on the Kolubara line if not W. of it. That is, from deployment to the first great battle the campaign would have to be carried out with road transport, and in the case of the rightmost routes (axis Valyevo-Chachak) pack transport only.
From the Middle Drina (Lyeshnitsa-Lyuboviya front) all routes converge on Valyevo, at the head of the Kolubara. Thence operations towards Arangyelovats, Gorni Milanovats and Chachak would take the course already discussed. But up to Valyevo operations would lie wholly in mountainous country * and would depend for supply almost entirely on pack transport from distant railheads for in Bosnia the railways come to an end 25-27 m. W. of the Drina, and from that river to Valyevo is a further 32 miles. All arguments against the choice of the Shabats route therefore applied with far greater force to the Valyevo route. Farther S., in the region of Vishegrad, operations into Old Serbia by way of Uzhitse might gain a footing, or help a Valyevo force to gain a footing, at Chachak. But such opera- tions would be mountain warfare pure and simple. Although in this part the railway runs right up to the frontier, no body of troops large enough to deal with the Serbian army could be col- lected and fed in the upper valley of the western Morava, either by way of Uzhitse or by way of Valyevo. As a threat to the flank and rear of a beaten Serbian army retreating up the Morava valley, such manceuvres might be effective; but to the beating of that army they could contribute practically nothing. Opera- tions from Vishegrad into New Serbia would be too eccentric to have the slightest influence upon the conflict of armies in the Morava valley, and the same would apply to operations from Focha or from the Herzegovina against Montenegro, supposing that state to join Serbia in resisting the Austrian advance. Operations in this quarter would be scarcely more than guerrilla.
On a review of the conditions, then, the most logical plan of campaign from the Austrian point of view would be a frontal operation S. from the Danube front, coupled with an enveloping movement from the Sava (or Sava-Lower Drina) front directed on Arangyelovats, the latter either preceding the other by some days and seeking to surprise possession of the Kolubara ridge, or following it and directed to the flank and rear of an ascertained position of the enemy. Loosely connected with these main opera- tions an advance from the Middle Drina or Vishegrad or both on Chachak and the western Morava had advantages for the exploi- tation of victory, but not for the winning of it. Such a plan was proposed before the war in the Austrian staff, and practically the same plan was assumed by the Serbians as the basis of their defensive deployment. Whether any or all of these operations could be carried through before the Russians intervened, or before their intervention became dangerous, would depend on available time. If no time at all were available, a pure defensive was the only course. If time for a thorough conquest, all the operations above mentioned with proper proportioning of strength would contribute something to a decisive success. In any intermediate conditions, one or more of the subsidiary opera- tions would be omitted from the plan, and, especially, if a mini- mum time were available only the direct S. attack from the Danube front would be admissible, with or without a collateral attempt by light forces from Shabats or Obrenovats to. seize the Kolubara ridge.
The calculation of this available time depended principally upon the arrangements made between the Austrian and the Ger- man staffs for initial operations in Poland, Galicia and E. Prus- sia. For this a plan had been drawn up by Conrad von Hotzen- dorf and Moltke in consultation; and, on the basis of this plan,
- Even the Yadar valley road was in many places unsuitable for
wheeled transport.
