When you have in this way both barrels flattened as nearly alike as possible and as straight as can be done by testing with a straight-edge, lay both barrels on a level surface, and see that the flattened places touch each other true and evenly. To know if the flattening has been done parallel with the outside flats at the breech (supposing these to be left on the barrels) place a small square on the same surface on which the barrels are laid, letting the upright arm of the square just touch the outside flat. If the square touch the flat alike from top to bottom, then the flats are parallel, but if there be an open space to be seen, then file away the flat to be joined until the square indicates that both inner and outer flats are parallel. Be particular in regard to this, as it is easier to correct the inner flattened surface than to make parallel the outer one. If the same amount be taken from the breech as from the muzzle, the point of divergence will not be sufficient for accurate shooting. More must be taken from the breech to allow of this divergence to be extended a greater distance. The flattening at the muzzle does not require to be very great, yet must be done to some extent; but, as before remarked, this will depend upon the smallness or size of the middle portions of the barrel. It is very doubtful if any two gunsmiths will agree as to the inclination of a pair of barrels. Perhaps it would be almost foolishness to insist on any definite or certain inclination, so that the converging lines would come to a point. As an instance, let the point of divergence be two and a half yards, it follows that at forty yards the right barrel, if the gun were rigidly held as if fixed in a vice, would throw its charge about six inches to the left of the mark, and the left barrel, _vice versa_. Let it be supposed that two barrels are each thirty-two inches in length and one-sixteenth inch thick at the muzzle and three-sixteenths thick at the breech, it requires the difference four-sixteenths be multiplied by the times the length of the barrels are contained in the forty yards, this being forty-five, to ascertain what distance the point of the different lines are from each other. In this case it is eleven and one-quarter inches, or five inches and five-eighths of an inch from the centre or line of sight. It may be remarked that if the point of convergence be forty yards it will be productive of as good results, perhaps, as any other.