Returning from this digression to the subject of recoil, we found that sandbags placed at a certain distance in rear of each wheel not only effectually checked the carriage, but also (a great consideration) ran it out again. This system was used both by the 4.7's and ourselves at the end of the war; and seeing that the guns had only half crews, it was a _most_ important saving to men who had perhaps marched ten miles, loaded and off-loaded ammunition, and then had perhaps to fight the guns under a hot sun for hours. To fill and carry the bags, however, is a nuisance, and some better system on the same principle is needed, such as the inclined wedges that I saw by photos the Boers were using in rear of wheels; and I should very much like to see some such system substituted for our present one. I have not seen the hydraulic spade used, perhaps that is the _best_. To put it briefly, the hastily improvised gun-carriage of the 12-pounders had, on account of this very haste, the following defects:-- (1.) Too weak generally in all parts, particularly wheels and axles, for any long campaign. (2.) Wheels and axles being a scratch lot, none in any of the batteries were interchangeable, which caused many times later in the campaign when wheels began to give out, much anxiety. Several times we only had guns ready for action or trekking by the "skin of one's teeth," and it must be borne in mind that any new wheels wired-for sometimes took two months to arrive on the very overcrowded railway--a single line. (3.) The system of checking the recoil of the field carriage was a bad one. (4.) All the 12-pounders except two were in the first instance sent up without limbers, and therefore had to be limbered up to wagons. This for practical purposes in the country we had to trek over was absolutely useless and caused endless delays. Eventually we all got limbers built at Maritzburg, and equivalent gun-oxen to drag the guns separately from the wagons. (5.) The trail of the gun consisted of a solid block of wood some 12 feet long; so that if one laid the gun to any long range (in most over 7,000 yards, I think) the oil cylinder under the gun, on trying to elevate it, would bring-up against this trail and prevent laying. This therefore necessitated digging pits for trails to shoot much over 7,000 yards, which in bad ground often took some considerable time. To obviate this defect would of course be very easy with a steel trail of two side plates, and space for gun and the cylinder between the sides.