This remark applies even to battles like Crecy or Agincourt, where the numbers slain were unusually high, and where they are said to have been accurately ascertained by counting after the victory. When Froissart on such authority quotes 1,291 as the total number of warriors of knightly or higher rank slain at Crecy, it is possible of course that he is not the victim of deception; but what of the 30,000 common soldiers for whose death he also vouches? A monk of St. Albans, also a contemporary, speaks only of an unknown number (_et vulgus cujus numerus ignoratur_); which in the account of the Abbot Hugo was put definitely at more than 100,000. It is evident from this that the greatest laxity prevailed in reference to chronicling the numbers of the slain; so that if we take 3,000 instead of 30,000 as the sum total of common soldiers slain at Crecy, it is probable that we shall be nearer the truth than if we implicitly accept Froissart’s statement. The same scepticism will of course hold good of the battles of the ancient world. Is it likely, for instance, that in a battle in which the Romans are said only to have lost 100 men, the Macedonians should have lost 20,000?[10] Or again, is it possible, considering the difficulty of the commissariat of a large army, even in our own days of trains and telegraphs and improved agriculture, that Marius in one battle can have slain 200,000 Teutons, and taken 90,000 prisoners? But whilst no conclusion is possible but that the figures of the older histories are altogether too untrustworthy to afford any basis for comparison, the calculation rests on something more like fair evidence, that in the fortnight between August 4, 1870, the date of the battle of Wissembourg, and August 18, that of Gravelotte, including the battles of Woerth and Forbach on August 6, of Courcelles on the 14th, and of Vionville on the 16th more than 100,000 French and Germans met their death on the battle-field, to say nothing of those who perished afterwards in agonies in the hospitals. Recent wars have been undoubtedly shorter than they often were in olden times, but their brevity is founded on no reason that can ensure its recurrence: nor, if 100,000 are to be miserably cast out of existence, is the gain so very great, if the task, instead of being spread over a number of years, requires only a fortnight for its accomplishment.