We will suppose the _Face_ of the _Punch_ is _Sunk_ flat and straight down into the _Matrice_; but yet it is a little too deep _Sunk_. Therefore he considers how much it is too deep: If it be but a little too deep, perhaps when the _Face_ of the _Matrice_ shall be made exactly flat (for yet it is but _Rough-Filed_) it may be wrought down to be just of an _Height against Paper_. But if the _Punch_ be _Sunk_ so much too deep that the smoothing the flat of the _Face_ on the _Using-File_ will not work it low enough; then with a _Bastard-cut flat-File_, he takes off (according to his Discretion) so much _Copper_ from the _Face_ of the _Matrice_ as will make it so much nearer as he thinks it wants to the _Face_ of the Letter. But yet considers that the _Face_ of the _Matrice_ is yet to smoothen on the _Using-File_, and therefore he is careful not to take too much off the _Face_ of the _Matrice_ with the _Rough-File_. He is also very careful that when he is to _File_ upon the _Face_ of the _Matrice_, to _Screw_ the _Face_ of it Horizontally flat in the _Vice_: And that in _Filing_ upon it, he keeps his _File_ directly Horizontal, as was shewed, _Numb._ 1. _Fol._ 15, 16. _Vol._ 1. For if he let his right or left hand dip, the _File_ will in its Natural Progress take too much off the side it dips upon, and consequently the _Face_ of the Letter on that side will lie shallower from the _Face_ of the _Matrice_ then it will on the opposite side. The like caution he makes, in _Filing_ between the _Top_ and _Bottom_ of the _Matrice_ on the _Face_. For if he _Files_ away too much _Copper_ toward the _Top_ or _Bottom_, the _Face_ of the _Letter_ on its _Top_ or _Bottom-Line_, will lie on that end shallower from the _Face_ of the _Matrice_. Then he considers by his _Proof-Letters_ how much too thick the right or left-side of the _Matrice_ is.