(_b_) Numbers, when they are quoted in connexion with a really ascertainable fact, and with regard to a precise and concrete circumstance, are nearly always reliable; though their significance differs, as I shall show in a moment, very greatly according to the way they are treated. Thus, if a Government says, “in such-and-such a place or on such-and-such a day we took three thousand prisoners,” it is presumably telling the truth, for the enemy who has lost those prisoners knows it as well as they do. But estimates of what has happened in the way of numbers, where the Government issuing the estimate can have no direct knowledge, are quite another matter. These are only gathered from prisoners or from spies, and are often ridiculously wrong. (_c_) All official communiqués of whatever Government conceal reverses, save in minor points. They are wise to do this because there is no need to tell the enemy more than he may know of his own success. Reverses are not actually denied. They are omitted. Witness all omission of Lemberg from Austrian or German communiqués and, until somewhat late, of Tannenberg in Russian, of Metz in French official accounts. Those are the three points which all the official communiqués have in common, and by bearing them well in mind we can often frame an accurate picture, in spite of the apparent contradiction and confusion which the reading of several communiqués one after the other produces. For instance, the Germans are trying to cross the Bzura River according to the Russian communiqué of Saturday. Next Wednesday the Russian communiqué says, “Two attempts to cross the Bzura at such-and-such places were repelled”; while the German communication says, “Our troops succeeded in crossing the Bzura River at such-and-such a village and established themselves upon the right bank.” In such a case the reader will be wise to believe the German communiqué and to take it for granted that while the Russians have repelled certain other attempts of the enemy to cross, this attempt has succeeded. But if the Germans go on to say, “The Russians retired after suffering losses which cannot have been less than twenty thousand,” that is no news at all. It is obviously conjecture. The various Governments issuing the communiqués have acquired certain habits in them which are worth noting if one is attempting to get at an accurate view of the war, and these habits may be briefly described as follows: