the heavy and dull methods which had hitherto controlled the action.  While Buller was talking about his tedious railway across the veld, and Milner at Capetown was dismalling the situation and discouraging the advance, Lord Roberts had in effect entered the capital of the Free State and seemed to have completed half his task.  The Boers were hypnotized and deceived not only by signs from which they drew wrong inferences, but also by bogus orders which it was arranged should come under their notice and which were simultaneously cancelled in cipher:  and when too late they awoke from the bewilderment, they began to scuttle to and fro like rabbits in a warren.  There is good reason to believe that if the strategic ability of Lord Roberts could have been united in one mind to the determination of Lord Kitchener the war would have been over in a year.

On February 8 Lord Roberts arrived at Modder River, where he found bad news awaiting him.  Buller had failed at Vaalkrantz, and the diamond men of Kimberley were threatening to capitulate.  By February 13 30,000 combatants, some of whom in order to preserve the illusion had been kept in the centre until the last moment, were in readiness at various points between the Orange and the Modder.  The immediate problem before Lord Roberts was the relief of Kimberley in combination with the cornering of Cronje.  In the background was the Natal trouble.  Buller was again helplessly wringing his hands and reaching round to find excuses for his misadventures.  Lord Roberts wisely left him alone and went on with his own work.  He saw what Buller refused to see, that the Tugela could be crossed at Magersfontein and Ladysmith relieved at a drift of the Modder River.

[Illustration:  Sketch map of the Riet and Modder Drifts.]

On February 11 Lord Roberts set his army in motion; and the operations of the next few days may be summarised with sufficient accuracy as a cavalry raid northwards, but avoiding Cronje’s left flank at Brown’s Drift, to relieve Kimberley; combined with an infantry advance to cut him off.  It was not possible to make the initial movements in the direction of the eventual advance, as the Magersfontein-Brown’s Drift quadrant N.E. of Modder River was strongly held by the enemy, and disallowed a cavalry advance from below the junction of the Riet and the Modder in the direction of Kimberley except by a westerly detour which could not be accommodated to the general scheme.  In order to strike the practicable drifts on the two rivers above their confluence, it was necessary for the advance to be made along the curve of a parabola which issued from Modder River Station in a S.E. direction, and in a sixty-mile circuit crossed the rivers and finally approached Kimberley, only twenty miles distant from the starting point, almost in the opposite direction.