Had the Japanese got possession of our communications, a catastrophe unprecedented in military history might have resulted. Without any victory in the field, the mere destruction of the railway in our rear, combined with the cutting off of local resources, would have threatened us with starvation—and disaster. Such were the unfavourable conditions under which we fought for fifteen months, and our army was not only not completely defeated, but grew in strength, while our communication with Russia gradually became better secured and more efficient. We had always recognized the possibility of being driven back to Harbin and beyond; but this never happened, and we held on to Hsi-ping-kai. The situation could only have been improved in one way—by a rapid concentration of sufficient troops for, and an assumption of, an offensive all along the line. While these troops were collecting, each fight—quite independent of its actual result—would have really helped us if it had at all weakened the enemy. But our departure from our accepted plan of operations began at the commencement of the war, when, instead of fighting a rearguard action, General Zasulitch got seriously engaged against the whole of Kuroki’s army at the Ya-lu, and was defeated.
In May, when the 3rd Siberian Division[12] had alone arrived at Liao-yang (besides the troops of the Pri-Amur Military District), the Viceroy, fearing for the fate of Port Arthur, instructed me to assume the offensive towards the Ya-lu against Kuroki’s army, or southwards for the relief of the fortress of Port Arthur. But the inadequate force with which General Shtakelberg pushed forward, owing to ignorance of the fact that the Japanese were in superior strength, got drawn into a serious engagement at Te-li-ssu, and was defeated. With the arrival of all the units of the 4th Siberian Corps and one division of the 10th Army Corps, it seemed possible to contain Kuroki’s army, to concentrate fifty to sixty battalions rapidly in the direction of Ta-shih-chiao, and to attempt to hurl back Oku to the south. It seemed as if our army had a splendid chance of operating on interior lines. The enemy was strung along three lines of advance—Dalny, Kai-ping, Ta-shih-chiao (Oku); Ta-ku-shan, Hsiu-yen, Ta-ling, Hai-cheng (Nodzu); Ya-lu, Feng-huang-cheng, Fen-shui-ling, Liao-yang (Kuroki). We occupied the central position—Liao-yang, Hai-cheng, Ta-shih-chiao—with advance guards thrown forward on to the Fen-shui-ling heights. We might have been able, by containing two armies and deceiving the enemy by a demonstration, to strike the third army in force. A blow delivered at Kuroki or Nodzu did not promise success, owing to our lack of training in, and unpreparedness for, hill warfare [we had no mountain artillery, our baggage was heavy, and we were uncertain of receiving supplies, owing to the insufficiency of transport material]. The only other course was to strike at Oku, who was based on the railway, but such an operation was risky, because Kuroki and Nodzu might have driven back our screens and fallen on our communications. On June 26 and 27, when only one brigade of the 31st Division of the 10th Army Corps[13] had arrived at Liao-yang, the Japanese on the eastern front (Kuroki and Nodzu) themselves took the offensive and seized the passes (Fen-shui Ling, Mo-du Ling, Da Ling) on the Fen-shui-ling heights. We opposed them in insufficient strength, and did not even make them disclose their numbers. The troops of the eastern force withdrew towards Tkhavuop, and General Levestam’s force to Hsi-mu-cheng. Our screens were thus situated as follows: on Kuroki’s line of advance, only two marches from Hai-cheng; on Oku’s line of advance at Ta-shih-chiao, four marches from Liao-yang.[14] Our position was critical, particularly if the information we had received as to the Japanese collecting in considerable force to operate against Hai-cheng was confirmed. Still, if we were able to strike a rapid blow at Oku, we might rob the enemy of the initiative, and after forcing back Oku’s army, have fallen on Nodzu. After we had driven back these troops, Kuroki’s position would have been so far forward and so far separated from the other groups that the danger of his breaking through to Liao-yang would have been minimized. But for such decisive operations the first requisite was the concentration of sufficient troops for offensive operations against Oku.
GENERAL BARON KITEN NOGI.
At the end of June we had altogether available against the three Japanese armies 120 battalions, and were inferior to the enemy both in the number of battalions and the number of men. Our position was made worse by an epidemic of dysentery which broke out amongst the troops at Ta-shih-chiao, and swept off a considerable number of men. The Krasnoyarsk Regiment[15] was the greatest sufferer, having as many as 1,500 men down with the disease at the end of the month. But the main thing which delayed any advance on our part was the rain, which made all moves difficult, and some places absolutely impassable for transport. It was even difficult to convey supplies to our various stationary forces over distances of less than a march. In spite of the lack of pack-saddles, wheeled transport had to be given up for pack transport, and not even pack-animals could do more than seven to eleven miles in the twenty-four hours. On the Liao-yang–Lang-tzu-shan road things were still worse, for the bridges over the mountain streams had been carried away, and communication between the eastern force (3rd Siberian Corps, under Count Keller) and Liao-yang was interrupted for some time. Far, therefore, from being ready to advance, the officers commanding the 1st and 4th Siberian Corps found the greatest difficulty in rationing their troops, and on June 29 asked that they might be withdrawn towards the positions near the railway at Ta-shih-chiao, and that the country east of the line might be left to the cavalry, with a few infantry units in support.[16]
General Count Keller was persistent in his demands that communication should be maintained between his force and Liao-yang, but we had neither the material, the means, nor the time to comply with his wishes, which would have meant the laying of a light railway and the strengthening of the road bridges. As I feared that the Japanese might make a fresh forward movement on Hai-cheng, I ordered thirty-nine battalions to concentrate near Hsi-mu-cheng on June 29. The short march from Hai-cheng was accomplished on the 28th with great difficulty through a sea of mud, and on the 29th Hsi-mu-cheng was temporarily cut off by the mountain streams in flood. The feeding of the troops collected there was found to be so difficult that as soon as it was known that the enemy, instead of advancing, had retired towards the Fen-shui Ling (Pass), certain units were ordered to return to the railway. Taking advantage, on July 18, of the screen formed by a portion of the 17th Army Corps, we attempted to advance against part of Kuroki’s army in the hope of forcing our way forward and gaining a partial success. For this Count Keller had under his command forty-three battalions, but the attempt failed. He stopped the action before any large number of our troops had become engaged. On the 29th Oku’s army took the offensive; we had to evacuate Ta-shih-chiao and Newchuang after a feeble resistance, and allowed Oku and Nodzu to join hands. When on July 23 I inspected the units of the 10th Army Corps, who were holding the position near Hu-chia-tzu, I found out how absolutely incapable of operating in hilly country the troops newly arrived from Russia were. Before sending them forward, it was necessary to train them in hill fighting, and to provide them with pack transport. On July 31 all three Japanese armies advanced, and we concentrated after a series of battles round Liao-yang. Here, in spite of our resistance, the three armies were able to join hands. Their attacks on the left bank of the Tai-tzu Ho were repulsed, but owing to the unfortunate nature of our operations on the right bank, the conditions became so unfavourable to us that I was obliged to order a retirement to Mukden. The withdrawal was conducted without the loss of a single gun or transport cart, while the enemy lost in men more heavily than we did. In the detailed accounts I have given in the first three volumes of the operations at Liao-yang, on the Sha Ho and at Mukden, our difficulties and the causes of our defeats are explained. The course of events showed that our original scheme of operations was quite a correct forecast, for in it the probable necessity of retiring towards Harbin had been foreseen. Indeed, matters at Liao-yang, on the Sha Ho, and especially at Mukden, might have been very much worse for us than they were, and might have necessitated our retirement on Harbin early in October, 1904, when, as a matter of fact, we remained in Southern Manchuria.
Clausewitz has truly laid down that an army should be inseparably connected with its base, but our base was Russia, more than 5,000 miles away. The way that this one difficulty alone was overcome will perhaps be eventually appreciated at its true worth. The very complicated attendant circumstances demanded great and patient efforts on the part of the whole nation in order to turn them to our advantage. Our reverses were explicable, and even in our defeat we exhausted our enemy, while ourselves increasing in strength. It was inevitable that a different complexion would have been put on the face of things as soon as circumstances became more favourable to us.
Difficulties in Organization.
The war showed that our army organization gave us too small a percentage of actual combatants as compared with the total numbers whom we rationed. By this I mean that, in spite of the immense numbers that we maintained in the face of great difficulties, we were unable to put enough men into action to win. Our establishments of all arms, of parks, hospitals, transport corps, field bakeries, staffs, and all offices and institutions, include a large percentage of non-combatants, which was swollen in the last war by the absence of any organized line of communication troops, the necessity of carrying out a large amount of railway construction, and of appointing officers and men to newly formed supply and transport units. Even so the number of non-combatants laid down in the establishments for each unit was not sufficient to perform the duties that fell to them, and it became necessary, for reasons which will be mentioned later, to detail combatants for domestic duties. As but few non-combatants were wounded in action, the proportion of them to the combatant element became still greater after every big fight. It was usual, when a battle was imminent, to order back to their units all men who were on extra-regimental duties, but in spite of all the steps taken, the fighting number was never more than 75 per cent. of the number of men on the strength. In the beginning of April, 1905, when we were preparing the theatre of war up to the River Sungari, the combatant element of the 1st Manchurian Army actually fell to 58 per cent. of the strength. As in previous wars, the infantry, of course, did most of the fighting, and also carried out by far the greater number of fatigues and extra duties. As they also lost more men in action, their fighting strength was proportionately more reduced than that of the other arms.[17] In April, 1905, the percentage of rifles in the 1st Manchurian Army to the total number of men that had to be rationed was 51·9 per cent. When the convalescents returned to the ranks, its strength amounted by the beginning of December to 192,000 men, of whom 105,879 carried rifles; but we could only put a much smaller number in action owing to various duties, fatigues, etc. In August, 1905, the number of rifles was 58·9 per cent. of the total of men rationed.
To obviate this state of affairs, and to insure that companies should be as strong as possible in action, I gave orders on June 9, 1905 [when I was commanding the 1st Manchurian Army], that out of each of the four battalion regiments, not more than 369 combatants should be detailed for extra duties. This figure included 128 stretcher-bearers, 35 bandsmen, and 48 men for baggage guards. In addition to this, a large number of men were required for road and bridge work on the communications, for guards for the different stores, for working parties to assist the supply and medical services, for policing villages, for duty with the improvised transport units, etc. True, this had its compensations, for we were able thus to get rid of the 2nd Category reservists from the ranks; but we felt the loss in the number of rifles we could place in the firing-line. Of course, there were, in addition, the sick, the wounded, and the convalescents with units and in hospital. In this way the total of all ranks classed as combatants but absent from the firing-line, or not doing combatant work, amounted on the average to 800 men out of every four-battalion regiment, or about one-quarter of its strength. To carry on the campaign without properly organized units on the communications, without sufficient camp guards, without making roads and bridges, without allowing men for transport and baggage duties, was impossible. Notwithstanding the good payment we offered, the native population did not come forward to work freely, especially when fighting was imminent. A certain number were employed on transport, but they were very unreliable, and bolted at the first alarm, often taking their horses and carts with them. During the battle of Mukden, for instance, the whole of the hired transport of the 1st Army, consisting of 400 carts, entirely disappeared. Our attempts to obtain Russian hired labour were a failure, though the rates of pay offered were liberal enough.