SOME RESULTS OF SECTION OF PARTICULAR NERVE BRANCHES TO THE AMPULLÆ OF THE FOUR VERTICAL SEMICIRCULAR CANALS OF THE FROG

1. A new method of ablation of each canal, namely, severance of its ampullary nerve external to the membranous labyrinth, has been used throughout. 2. With the group of four vertical canals all possible ablatory selections and combinations have been effected. In the case of the anterior ampulla it has been possible to examine the effect of partial ablations, one half of the crista having been rendered functionless in some animals, the other half in others. The success of these various operations has been checked both by gross and by microscopic examination. 3. Removal of any single vertical canal documents itself in some disturbance of behaviour during ordinary quiet forward crawling. The particular single lesion can be diagnosed by observation of the crawling gait alone. It is similarly revealed when the frog is induced to circle on a revolving turn‐table. 4. Sudden tipping of the frog towards the body corner corresponding to an ablated ampulla causes extension of both limbs on the side opposite to the functionless canal, with concomitant active flexion of one (or both) limbs on the side of this canal. It is especially the positive extensor thrust of the limb on the corner diagonally opposite to the functionless canal that then tends to cause overbalancing. The animal does not simply fall over, it is partly pushed over by this unusual action of its own limbs. 5. Ablation of a diagonal pair of canals produces unusually severe disturbance of gait. On attempted movement the frog becomes a prey to heavy bodily swing of a pendular or regularly periodic nature. The pendulation occurs about a diagonal horizontal axis and in a plane whose direction corresponds with that of the pair of obliterated canals. In this diagonal plane the stability of the animal is obviously far from secure. At each extreme phase of the pendular swing two limbs of one side are in extension and the opposite two in flexion. Because of the regular alternation between these modes of limb innervation, an alternation that coincides with a diagonal downthrust of the head at one phase and with a diagonal uptilt of the head at the other phase, the disturbance in question is very striking. Even after the animal has ceased to progress, the pendulation does not immediately die down; the instability continues to manifest itself for a little after progressive effort has stopped. 6. The animal with a diagonal ablation of canals is unstable not only during progression but also when it is tipped on a tilt‐table. To sudden tipping displacement of the body in the diagonal plane corresponding to the pair of intact canals it reacts as promptly and precisely as a normal animal. Tipped, whether forwards or backwards, in the other diagonal plane, it is conspicuously unstable and is readily made to overbalance. Even slow tilting of the animal in this plane is apt to bring on pendular oscillation. A similar diagonal pendulation, coinciding with alternate beat of the hind‐limbs, may occur during the act of swimming. 7. One of the outstanding functions of the semicircular canals is instantaneously to check exaggerated rotational displacements of the head, which, but for the equipment of canals, are apt to accompany any, the slightest, movement of the body as a whole. 8. CRUM BROWN'S idea that each member of a functionally associated pair of canals responds to rotation in one particular sense and not to rotation in the opposite sense, has been fully substantiated. In the case of the frog, at least, this singularity constitutes a very important and fundamental feature of canal propriation. 9. While the canals are directly connected with the musculature of all four quarters of the body, one of the prominent servations of any single vertical canal is to act with unilateral emphasis upon that field of musculature which modifies orientation of the head. In some animals this may lie largely in the neck. In the short‐necked frog it includes fore‐limb and fore‐body muscles as well. If, according to their action on the head, one divides these muscles into four groups—(1) elevators of the right side of the head, (2) elevators of the left, (3) depressors of the right, and (4) depressors of the left—then (overlooking for the moment the possibility of reciprocal innervation or other composite interaction) the right anterior canal promptly activates (1), the left anterior canal activates (2), the left posterior canal activates (3), and the right posterior canal as promptly activates (4). By this means the original orientation of the head tends to be less disturbed during any active or passively imposed body movement. 10. Two simple tests, the “chin‐lift” and the “chin‐drop” test respectively, are described. These head tests, which require no tilt‐table, enable one speedily to recognise any mere uncomplicated lesion of one or more of the individual vertical canals, and in each case to specify the particular canal or canals involved. 11. Any lesion of a single vertical canal is apt to affect the ultimate stationary posture of the frog. A diagonal deviation from symmetry, in each case distinctive to the particular canal ablated, may then on occasion be observed. These deviations are in no sense “forced positions.” They are mere residual postures, which stay behind and exhibit a body‐taxis that during previous movement had been allowed temporarily to prevail, because the absent canal had failed at a critical juncture to function. These decanaliculate residual postures are all well within the range of asymmetrical postures assumed on occasion, or at need, by the intact animal. At any moment the operated frog may spontaneously correct a decanaliculate residual pose. 12. After a double anterior ablation the tone of the extensor muscle of the neck of the frog is defective. After double posterior ablation the tone of the flexors of the neck is similarly impaired. 13. That variety of labyrinthine “tonus” to which EWALD directed attention (without, however, discriminating it from another form of labyrinthine tone) as affecting the capacity of the relevant muscles “to act quickly,” proves to be of canal derivation. The receptors respond to sudden rotational displacements of the head. The tone depends upon or consists of simple, flick‐like, muscle pulls applied at the proper moment. These flicks, in respect of any single muscle affected, are precisely similar to tendon reflexes. 14. There is this difference between canal (tone) reflexes and tendon reflexes. The latter are strictly local responses, the former generalised responses, affecting muscles in all quarters of the body. The two innervational systems may work in conjunction with one another. Thus, in dropping from a slight height, the local reflex may affect individual muscles of the limbs subjected to impact. Meantime the canal reflex provides for co‐ordinated action both in these limbs and in the rest of the body. The special path for the generalised canal reflex is probably the vestibulo‐spinal tract. 15. A slight downward tap administered on the head of an upsquatting frog with a double ablation of anterior canals brings the animal to the ground with a great forward slew. Similarly a tiny, sharp, upward impulse communicated to the head of a pronely postured frog with a double posterior ablation makes the animal's fore‐body to rise with a great backward slew. In either case the massive nature of the result seems quite disproportionate to the trivial cause. 16. Ablation of two anterior canals, of two posterior canals, or of two canals of the same side, does not give rise to pendular swinging movements of the body. The functional antagonist of one anterior canal is the diagonally opposite posterior canal, and vice versa. Only when a pair of antagonists is removed does pendular oscillation occur. 17. When two canals of the same side are rendered functionless, the more common type of resulting, residual pose is not strikingly different from that due to simple loss of an anterior canal. A disability peculiar to the frog with two canals of the same side ablated, is a tendency to land on the operated side after a leap. While it may keep its natural horizontal trim during swimming, it may also swim with the operated side depressed. 18. After ablation of any single member or of any group of the four vertical canals, the frog tends after a leap to land in unusual fashion. Instead of having the body‐taxis proper to the moment of landing (that, namely, of a symmetrically up‐squatting frog), it has one or other of the various forms of body‐taxis that may be imposed upon it by simple slow inclination of a tilt‐table. The aberration resides, not in that the taxis in question is unnatural to the animal, but in the fact that it is the wrong taxis for that particular momentary juncture. Observation has shown that the aberrant forms of body‐taxis are possibly imposed at the time when the operated frog first vigorously propels itself off the ground. Continuing in existence during the course through the air, they are there at the moment of landing. 19. When all four vertical canals are removed, the frog on any attempted movement is apt to show heavy pendular oscillation of the body. The pendulation may now occur about any horizontal axis, antero‐posterior, diagonal, or transverse. During swimming the commonest form of pendulation is that about an antero‐posterior axis. Then the animal, without, however, turning turtle, swings like a swaying boat, so that the white of its belly may show first on one side and then on the other. A large swing in one direction tends to induce a correspondingly large swing in the opposite direction. Small initial swings, on the other hand, tend to be followed by correspondingly small counterswings. In shallow water, because of foot support, the extent of lateral oscillation is less. The most striking form of pendulation is that about a transverse axis, the body of the animal travelling forward and backward with each sway, like that of a lumberman sawing wood. Throughout a series of pendular swings, the axis about which pendulation occurs does not necessarily stay fixed. When it varies, it varies continuously, like that of a compound pendulum. When the animal moves its limbs alternately, as it may be forced to do on a revolving turn‐table, the individual quadrants of the body heave up and down in a curiously tumultuous way. 20. Subjected to slow tilt on a tilt‐table, the frog with four vertical canals removed responds perfectly. In making each new correction, however, its head and body may exhibit minor pendulation. When quickly tipped on the tilt‐table the animal first makes a great slew in the appropriate direction, and then by second slew of opposite sense takes the pose suitable to the changed inclination of the table‐top. 21. When three vertical canals have been ablated, the animal is far less subject to variety of pendular sway than is the frog deprived of four vertical canals. It rather resembles the animal with an uncomplicated diagonal ablation, for pendular sway during any type of activity is now restricted to the diagonal plane corresponding to that in which two of the canals have been removed. In the type of lateral fall occasionally observed on landing after a leap, this frog resembles one with an ablation only of two canals of the same side. 22. In the frog deprived of three vertical canals three residual poses may appear. Two of these are the diagonal forward and the diagonal backward postures corresponding to the plane of its ablated diagonal pair. The third is either the symmetrical grovelling or the high symmetrical up‐squatting, according as the animal has lost a pair of anteriors or a pair of posteriors respectively. The animal does not readily show a residual pose indicative of loss of the canal whose antagonist still persists. 23. The nerve of supply to an anterior canal crista branches like a Y externally to the membranous ampulla. The one twig supplies one end of the cristal ridge, the other twig supplies the other. Whichever of the two branches has been individually cut, the result is, to all practical means of determination, the same. The defect resembles that due to transection of the whole nerve, but is less marked. 24. Whereas horizontal canal function and utricular function may be complicated by the fact of concomitant vision, it is doubtful whether concomitant vision affects muscular responses to vertical canal stimulation. Our grateful thanks are due to the American Academy of Ophthalmology and Otolaryngology, which, by providing W. J. McNALLY annually with a research fellowship during a period of years, has greatly contributed to the promotion of the work here reported.