We also found difficulties about the capturing of guns.  At first we had merely provided that a gun was captured when it was out of action and four men of the opposite force were within six inches of it, but we found a number of cases for which this rule was too vague.  A gun, for example, would be disabled and left with only three men within six inches; the enemy would then come up eight or ten strong within six inches on the other side, but not really reaching the gun.  At the next move the original possessor of the gun would bring up half a dozen men within six inches.  To whom did the gun belong?  By the original wording of our rule, it might be supposed to belong to the attack which had never really touched the gun yet, and they could claim to turn it upon its original side.  We had to meet a number of such cases.  We met them by requiring the capturing force—­or, to be precise, four men of it—­actually to pass the axle of the gun before it could be taken.

All sorts of odd little difficulties arose too, connected with the use of the guns as a shelter from fire, and very exact rules had to be made to avoid tilting the nose and raising the breech of a gun in order to use it as cover. . . .

We still found it difficult to introduce any imitation into our game of either retreat or the surrender of men not actually taken prisoners in a melee.  Both things were possible by the rules, but nobody did them because there was no inducement to do them.  Games were apt to end obstinately with the death or capture of the last man.  An inducement was needed.  This we contrived by playing not for the game but for points, scoring the result of each game and counting the points towards the decision of a campaign.  Our campaign was to our single game what a rubber is to a game of whist.  We made the end of a war 200, 300, or 400 or more points up, according to the number of games we wanted to play, and we scored a hundred for each battle won, and in addition 1 for each infantry-man, 1-1/2 for each cavalry-man, 10 for each gun, 1/2 for each man held prisoner by the enemy, and 1/2 for each prisoner held at the end of the game, subtracting what the antagonist scored by the same scale.  Thus, when he felt the battle was hopelessly lost, he had a direct inducement to retreat any guns he could still save and surrender any men who were under the fire of the victors’ guns and likely to be slaughtered, in order to minimise the score against him.  And an interest was given to a skilful retreat, in which the loser not only saved points for himself but inflicted losses upon the pursuing enemy.

At first we played the game from the outset, with each player’s force within sight of his antagonist; then we found it possible to hang a double curtain of casement cloth from a string stretched across the middle of the field, and we drew this back only after both sides had set out their men.  Without these curtains we found the first player was at a heavy disadvantage, because he displayed all his dispositions before his opponent set down his men.