The extreme inner edge of the cleft velum should be seized near the base of the half uvula with the catch forceps (Fig. 62). The narrow-bladed knife (Fig. 61C) is entered with the back towards the tongue, just in front of the forceps, and made to cut the merest shaving from the margin as far as the apex of the cleft. Before relaxing the grasp of the forceps, the same process is continued backwards to the apex of the half uvula. The other side of the cleft is similarly treated, and, if possible, the strip of marginal tissue removed should be continuous throughout, thus satisfactorily proving that the whole of the cleft has been pared. This strip should be cut square with the palate, for if bevelled, the edges cannot afterwards be brought into such accurate apposition. Care should be taken in this proceeding not to contravene the important canon of plastic surgery, that no unnecessary amount of tissue should be removed; for it is most important to remember that in these cases, there is no excess of material, and that a too free removal of marginal tissue will lead to increased tension in the united palate, and subsequently to a less satisfactory functional result from defective closure of the posterior nares. STAGE III.—_Passage and Tightening of Sutures._ The quickest method and the one calculated to disturb the parts the least is a modification of that introduced and practised by the late Sir William Fergusson, the so-called “loop-method.” It consists in the passage of a loop of fine silk through both sides of the cleft, to act as a carrier for the silver wire which is to be the permanent suture. One of the needles already described, previously threaded with a piece of fine silk about sixteen inches in length, so that its ends are equal, is passed from the buccal aspect through the loose flap close to the margin of the cleft (_i. e._ about 2 or 3 mm. from it), and as near as possible to its anterior extremity. To accomplish this it is unnecessary to hold the flap with forceps, as its margin may be seriously damaged. The needle track should be perpendicular to the palate surface, and therefore parallel to the pared margin of the cleft. The silk is then seized close to the eye of the needle with the smooth-nosed forceps introduced within the cleft, the needle withdrawn, and the loop pulled forwards sufficiently to be laid temporarily on the side of the cheek. The same process is repeated at an exactly corresponding point on the opposite side, so that now there are two loops emerging from behind forwards through the cleft. By loosely threading the right loop through the left and gently withdrawing the latter, the former is carried through the flap on the left side (Fig. 69); in this way we have a double thread, with its loop on the left side and its free ends on the right, passing through the flaps on either side. This process is repeated at intervals of about 5 to 6 mm. throughout the length of the cleft from before backwards,