Now, how were these tactics to be met? Roberts nearly always endeavoured to meet them by still wider extensions, designed to overlap the enemy’s front. He planned to throw substantial bodies of mounted troops right round one or both of the hostile flanks, with the view (as at Poplar Grove) of intercepting the enemy’s retreat. These movements never led to interception, though they were generally successful as turning movements which led to the enemy’s retreat—a very minor object. On the other hand, they were exhausting to horses and men alike, reducing offensive power when, after long riding, it was at last called for, to a point below the normal, and the normal was not nearly high enough.
Zand River (May 9 and 10) illustrates this class of action. There, 4,000 mounted men under French and Hutton on the left, and 3,000 under Broadwood and De Lisle on the right, were deputed to get round both flanks of a front of twenty-five miles, held by 8,000 Boers. French, having passed six miles outside the last Boer post on the 9th, got well round to the rear on the 10th, with his Cavalry leading and his mounted riflemen in support, but was then held up for several hours by small detachments, and suffered considerable loss. He covered thirty miles on the 10th, and could not, owing to the condition of his horses, respond on the same night to a suggestion by Roberts for raiding Kroonstad. Broadwood’s turning movement was abortive, partly through an accidental withdrawal of his horse battery, but mainly through the circumstance that the Boer left (wide as Hamilton’s extension was) still overlapped our right, and that the overlapping portion, not content to remain on the defensive, endeavoured during the morning to envelop our extreme right. Botha effected an orderly retreat, his centre maintaining a good show of resistance against the Infantry and Artillery attacks. With our main body there was a brigade of Cavalry and considerable numbers of mounted riflemen.
Diamond Hill, where Botha defended thirty miles of hills, was a still more extreme instance of the same method. French, with 1,400 Cavalry and mounted riflemen, was designed to ride right round the enemy’s right, and cut the railway in his rear—a ride of at least thirty-five miles, without any allowance for interruptions or détours. Broadwood, with 3,000 men, was to turn the enemy’s left and support our right attack. The centre was to be withheld until one or both of these movements should succeed. Botha had anticipated these tactics and had strengthened his flanks accordingly. Both mounted columns were held up, and stood for a time in considerable danger of envelopment. On the second day the centre was forced by Infantry, aided, and very effectively aided, by mounted riflemen.
It must be remarked that our total strength at Diamond Hill was unusually small—14,000 men in all, of whom 4,800 were mounted, and 64 guns. The Boers had 6,000 men and 20 guns.
Now, there is but one way of looking at situations of this sort. If we are seeking instruction for further wars, we must recognize that the only sound method of combating such prodigiously wide extensions of a numerically weak enemy is to force his line instead of turning it. To devote the major effort to turning it is to play into his hands, to permit him by sheer bluff to impose exhausting tactics which neutralize your own numerical superiority.[[47]] The difficulty was to apply forcing tactics against so formidable a foe as the Boers. Our crying need all along was tackling power with the horse and rifle combined—high, mobile tackling power, based on surprise and speed, and taking the form, where need be, of mounted charges into or through the enemy, on the lines afterwards taught us by the Boers, and already exhibited by them at Sannah’s Post. Again and again, in reviewing the South African combats, we look back to the Klip Drift charge of February 15, 1900, with profound regret that its true lessons were not laid to heart and its false lessons discarded. There was the germ of success. Add operative tackling power to the nerve required to ride through fire, eliminate the arme blanche and every last vestige of tactical theory connected with it; eliminate as far as possible Artillery preparation and support; be content with a reasonable superiority of strength, and there you have for future wars the true tactics of mounted offence.
It is impossible to blame Roberts for over-reliance on wide turning tactics. In the last resort, whatever the scheme employed, whether we rode wide or rode through, success depended on sheer fighting capacity in the ultimate fire-fight. Nothing could replace that. Roberts could only endeavour to make the best of the material to hand. His frequent attempts to encircle far-flung fronts were an instinctive recognition of inadequate aggressive power in his mounted troops. The prejudice, so general in South Africa, against “frontal attacks” by Infantry was often a reflection of the same instinct, that is, of an instinct to avoid heavy losses which could not, unaided, lead to a decisive result. In point of fact, all attacks eventually become frontal, in the local sense. And, in the case of mounted troops, it was of no avail to send round a large body of men to take the enemy in flank or rear, unless they were able to burst through frontally the detachments sent against them.
Still less tenable is the suggestion that the right course for Roberts was to have projected still vaster and more circuitous mounted operations, designed to cut the enemy’s communications far in rear of the zone of immediate hostilities. French is said to have favoured this course more than once, but did he realize what it involved? If the requisite speed were sustained, the horses, already tried to the limit of endurance, would have suffered from that very over-exhaustion of which there had been so much complaint in the past. But, in fact, such raids, on the scale of those made by Stuart, Wilson, and the Civil War leaders, entailed complete independence of the main army, an object never attained in South Africa without transport arrangements which reduced speed to too low a level. The question, of course, was not peculiarly a “Cavalry” question—for raids, American, South African, or Manchurian, turned exclusively on fire-action. I shall be compelled, nevertheless, to argue the matter again, in Chapter XII., on a “Cavalry” basis, taking Zand River once more as an illustration.
2. It must not be supposed that frontal or semi-frontal attacks were not tried by the mounted troops. Local circumstances often brought them about. Generally, however, they tended, even locally, to take a too circuitous form, the tendency, inevitably, being more noticeable among the Cavalry, with their inferior firearm, than among the mounted riflemen.
These latter troops, now possessing an acknowledged and independent status of their own, and led by some able men like Hutton, Alderson, and De Lisle, did remarkably well in some instances, though poorly in others. The Australians and New Zealanders seem always to have shown the most tactical vigour. Hutton’s fight on May 5 to secure the passage of the Vet on the left of the main army was a good performance. The mounted riflemen did well also in the pursuit north of Johannesburg on May 30, in the fighting outside Pretoria on June 5, at Diamond Hill on June 12, and on several other occasions.
French’s operations outside Johannesburg on May 28 and 29, when, prior to the arrival of the Infantry, both classes of mounted troops were employed in unison, are interesting. French was in his best mood. There was no lack of vigorous will on the spot, but the turning movements by the Cavalry (except the last, which followed the Infantry assaults), and the frontal attacks by both classes, alike failed. There would seem on this occasion to have been a good opportunity for a rush through the centre on the lines of Klip Drift.