g axis of the object
be made to correspond to that of the esophagus before traction for
removal is made (Fig. 92). Should the intruder be grasped in the
center and traction exerted, serious and perhaps fatal trauma might
ensue.

[191] [FIG. 92.--The problem of the horizontally transfixed foreign
body in the esophagus. The point, D, had caught as the bone, A, was
being swallowed. The end, E, was forced down to C, by food or by blind
attempts at pushing the bone downward. The wall, F, should be
laterally displaced to J, with the esophagoscope, permitting the
forceps to grasp the end, M, of the bone. Traction in the direction of
the dart will disimpact the bone and permit it to rotate. The rotation
forceps are used as at K.]

[FIG. 93.--Solution of the mechanical problem of the broad foreign
body having a sharp point by version. If withdrawn with plain forceps
as applied at A, the point B, will rip open the esophageal wall. If
grasped at C, the point, D, will rotate in the direction of F and will
trail harmlessly. To permit this version the rotation forceps are used
as at H. On this principle flat foreign bodies with jagged or rough
parts are so turned that the potentially traumatizing parts trail
during withdrawal.]

The extraction of broad, flat foreign bodies having a sharp point or a
rough place on part of their periphery is best accomplished by the
method of rotation as shown in Fig. 93.

_Extraction of Open Safety-pins from the Esophagus_.--An open safety
pin with the point down offers no particular mechanical difficulty in
removal. Great care must be exercised, however, that it be not
overridden or pushed upon, as either accident might result in
perforation of the esophagus by the pin point. The coiled spring is to
be sought, and when found, seized with the rotation forceps and the
pin thus drawn into the esophagoscope to effect closure. An open
safety-pin lodged point upward in the esophagus is one of the most
difficult and dangerous problems. A roentgenogram should always be
made in the plane showing the widest spread of the pin. It is to be
remembered that the endoscopist can see but one portion of the pin at
a time (except in cases of very small safety-pins) and that if he
grasps the part first showing, which is almost invariably the keeper,
fatal trauma will surely be inflicted when traction is made. It may be
best to close the safety pin with the safety-pin closer, as
illustrated in Fig. 37. For this pur