My first three volumes supply the answer to the most serious of these accusations: in them are described the tremendous efforts we had to make to prevent our operations ending worse than they did. I have never been one of those who believe that an order once given should not be countermanded or modified. In war circumstances change so quickly, and information received so frequently turns out to be false, that it would be fundamentally unsound to insist, in spite of changed conditions, on keeping exactly to an order once issued. An excellent example of this is given by the operations at Hei-kou-tai. The order received by the officer commanding the 1st Siberians to rest his troops on January 27, and to occupy the line Hei-kou-tai–Su-ma-pu–Pei-tai-tzu, was founded on the incorrect supposition of the commander of the 2nd Manchurian Army that San-de-pu had been captured. The former was more than once told not to attack. Yet, even though news was received that San-de-pu had not been taken, he insisted in carrying out the orders given, in which, by a mistake, a village held in force by the enemy was appointed as our halting-place. The result is known: we fought all day, lost 7,000 men, and at daybreak on January 28 were compelled to retire. With regard to the accusation that the late Commander-in-Chief[26] constantly countermanded his own orders, it is interesting to note that General Grippenberg, in his article, “The Truth about the Battle of Hei-kou-tai,” points out that, although he did not agree with him as to the necessity for retiring the right flank of the 2nd Army to take up a more concentrated position, he did not express this opinion to the Commander-in-Chief, because he and all his staff knew that Kuropatkin would never countermand an order once given.