sinus is situated for the greater part in front of the ganglion; hence, in enucleating the ganglion, hemorrhage from the sinus will be better anticipated if the former be pried from behind and pulled towards the outside.

A dangerous source of hemorrhage during gasserectomy is injury to the internal carotid artery. The artery lies immediately inside the body of the ganglion. It is intimately connected with the dura of the ganglion. To avoid hemorrhage from this vessel, it is well to carefully dissect away the ganglion with a blunt dissector and make sure that the dissector can be passed between the two structures before any tension whatever is put upon the nerve element. In many of the skulls examined by us, the carotid artery never came in contact with the ganglion, because the former ran all the way through a bony canal, in the apex of the petrosal bone; only when it emerges to form the ophthalmic artery did it come close to the ophthalmic division of the ganglion. So much for some of the mechanical difficulties in attacking the ganglion directly. Let us now direct our attention to the branches of the ganglion for the purpose of ascertaining how and where the nerve roots may be injected.

To continue our analogy of the Gasserian ganglion and its branches to a hand with three fingers, let us imagine that these fingers reach out through three foramina at the base of the skull to grasp certain anatomical structures. The first of these branches, known as the ophthalmic division, passes out through the sphenoidal fissure. (Fig. I, a.)

The second branch known as the superior maxillary leaves the skull through the foramen rotundum. (Fig. I, b.)

The third division known as the inferior maxillary leaves the skull through the foramen ovale. (Fig. I, c.) Through its own function, as derived from the Gasserian ganglion, this branch leaves the skull in company with the motor root of the fifth nerve and the two become intimately associated immediately they attain the pterygoid region. Thenceforth the inferior maxillary division of the Gasserian ganglion becomes a mixed nerve. Its motor fibres, supply the tensor tympani and palati muscles, as well as the muscles of mastication, the mylohyoid and anterior belly of the digastric.

Briefly speaking the fifth cranial nerve, by means of the Gasserian ganglion controls either altogether, or for the most part, the sensory innervation of the eye, the skin of the face, the teeth and the mucous membrane of the nose and mouth. One can readily perceive the enormity of reflexes which may be carried to the brain along this route. It is when these reflexes become painful that surgically we seek to inhibit the transmission of the particular impulse, or failing to do so, turn our attention to excision of the ganglion or section of its sensory root.

Before resorting to the radical operation, it

Let us follow the individual nerves as they leave the ganglion. We are struck immediately by two peculiar facts: Firstly, each division, though destined to innervate the exterior gives off a branch inside of the cranium, and secondly, each of these roots contributes a small branch to enter into the formation of new ganglia.

The other division of these ganglia is derived from the sympathetic system.

We sometimes meet with cases of trifacial neuralgia unbenefitted by surgical treatment of a peripheral branch. Might this not be due to the fact, that the intracranial branch is the one affected and that it has not been reached in the treatment?

The new ganglia bring the fifth nerve in communication with the sympathetic system-what an immense source for reflexes to be brought from distant parts to the trifacial nerve:

I am led by my observations to include another factor in the destination of the trigeminal branches-the fifth nerve is in some way or other connected with all of the other cranial nerves. We know this to be a fact with the third, fourth, sixth, seventh and ninth nerves. I cannot help but think that when the nasal branch of the ophthalmic division enters the nose through the nasal fissure, that it comes in contact with filaments of the olefactory nerve. It seems reasonable to believe that in the eye the ciliary branches of the ophthalmic root should come in contact with the terminal branches of the optic nerve. The auditory nerve sends a branch to the geniculate ganglion of the facial nerve, and the latter is considerably in communication with the branches of the fifth nerve. No great hardship would be encountered in tracing a connection between the pneumogastric and trifacial by means of the sympathetic system.

Confronted with a case of trifacial neuralgia and desirous of employing means towards treating the peripheral regions first, we find, that although on the face, the branches are here, there and everywhere, yet each main division makes its appearance through a distinct foramen and reigns over a definite area. If the terminal filaments of the three nerve trunks were traced over the face, they would form a mask which would cover it all but the ears and lower angles of the lower jaw.

The three bony foramina through which the nerve trunks reach the face are known respectively from above downwards as orbital, infraorbital and mental. The nerves passing through them have the same name as the foramen. These foramina all lie in a vertical line

draw downward from the supraorbital foramen. to the interval between the two bicuspid teeth. This line crosses both the infraorbital and mental foramina (Fig. II.).

With this preliminary consideration of the trifacial field in which only a few of the more important facts have been emphasized let us consider briefly the injection treatment of trifacial neuralgia and its practical application.

First: Let me call your attention to the fact that the three Gasserian branches have a deep exit from the cranium and a superficial exit (See Fig. 1). Their deep exit, let me recall, is from the sphenoidal fissure, the foramina rotundum and ovale, their superficial exit, the foramina supraorbital, infraorbital and mental. It is quite evident that the injection of the superficial branches as they exit on the face is a procedure of little difficulty, requiring but superficial knowledge, since the foramina are all palpable or the face and are not over 1⁄2-inch distant from the surface. The injection of these branches, therefore, need not concern this inquiry. But it has been found that the injection of the superficial branches is not always effectual, and it is the extension of this treatment to the injection of the branches as they exit from the foramina ovale, rotundum and sphenoidal fissure that presents a problem involving primarily anatomic accuracy. In other words, can a needle be passed with safety to the deep exit of these nerves; can the nerve be reached with a fair degree of certainty and what are the anatomical guides?

We have experimented on fifty skulls by exposing the foramina of exit within the skull, passing the needle to the point chosen and injecting methyl blue solution. When the methyl blue immediately appeared through the foramina we concluded that the needle point had reached the position desired.

We present, therefore, for your consideration the following guides for deep injections of the ophthalmic, and the superior and inferior maxillary divisions as they exit from the sphenoidal fissure, the foramina rotundum and oval respectively. The writer uses a needle 5 cm. long and 1.5 mm. in diameter, without stylet. To inject the nerve at its exit from the sphenoidal fissure insert the needle at the rim of the orbit at a point corresponding to the fronto-malar articulation. Pass the needle into a depth of 31⁄2 cm., being sure to hug the outer wall of the orbit. Special accuracy must be used to avoid the optic nerve. (Fig. III. a.) To inject the nerve at its exit from the foramen rotundum insert the needle at the lower border of the zygoma at a point opposite the posterior border of the frontal process of the malar bone. Push the needle directly inwards and slightly upwards for a distance of 5 cm., the point of the needle finally reaches the pterygo-maxillary fossa in which is found the nerve making its exit from the rotundum (Fig. III, b). If the needle after passing the subcutaneous tissue strikes bone, it is the coronoid

process of the lower jaw. The way must be cleared by having the patient open his mouth and placing a gag between the teeth.

To inject the inferior maxillary division at its exit from the foramen ovale, insert the needle at lower border of the zygoma 1⁄2 cm. in front of the condyle of the jaw, push the needle directly inward and slightly downward for a depth of 4 cm., at which distance the point of the needle is at the ovale foramen. (Fig. III, c.)

It may be well to state that the injection is seldom made directly into the nerve, but rather in the vicinity of the nerve and the diffusion of the agent accomplishes the desired purpose.

The solution used is the cocaine, chloroform, alcohol solution of Patrick & Hecht, consisting of a grain of cocaine, X minims of chloroform, and 2-oz. of 80% alcohol.

A curious fact about the alcohol injection is that it does not destroy the nerve, but in some way changes its chemistry. In other words, the pain is relieved, but sensation is not destroyed. The patient suffers considerable pain during the injection, but these patients are accustomed to excruciating pain and are willing to undergo temporary suffering if relief is the reward.

No preliminary anesthesia is used, for the reason that the best evidence of a successful puncture is the spasm of pain, numbness and anesthesia over the course of the nerve successively reported by the patient.

Ecchymosis and swelling are to be expected and trismus following injection of the third division is to be anticipated, but they are never more than a temporary inconvenience.

Some Practical Observations.

In the practical application of this treatment the author makes the following suggestions as the result of his clinical observations:

a. In all cases where no previous operation has been performed it will be well to first apply the injections to the superficial exits of the nerves as they appear on the face before making the deep injections. In some of our cases such treatment seemed to be sufficient and there is no need for making the deep injections unless the pain returns.

b. In cases the subject of previous operation (gasserectomy, Abbe's operation, etc.) the deep injection only was effectual.

c. Although in the majority of cases the pain is almost immediately relieved, in a few instances the pain is worse after the injection and may continue for a few days. This condition is undoubtedly due to increased swelling of the parts in the vicinity of the nerve and subsides as the swelling decreases. In one of our cases intense pain followed deep injection of the second division and continued for seven days, after which the pain disappeared and the patient is entirely free from pain for the first time in ten years.

d. The solution used should be freshly prepared for each treatment. The ingredients undergo rapid chemical changes.