The chronic stage is marked by the development of a new set of symptoms—contractions of certain muscles surrounding one or more joints and deformed positions of the limbs. These symptoms do not always appear. If all the muscles surrounding a joint are completely paralyzed and extensively atrophied, and if no weight is imposed on the limb by the action of a non-paralyzed upper segment, and if the paralyzed segment be so supported that its own weight does not approximate the insertion-points of muscles, and thus cause their passive retraction, then there is no deformity, but a dangling limb, a membre de Polichinelle.

Laborde states that contractions appeared at the earliest about two months after the paralysis. Seeligmüller, however, has seen pes equinus and pes calcaneus develop in four weeks. The date is partly, at least, determined by the time at which the children try to walk or otherwise to use the paralyzed limbs; and the deformities are very much more marked in the lower extremities, proportioned to the much greater weight which they are obliged to sustain.

All varieties of club-foot, and most frequently equino-varus, knock-knees, rigid flexions at the knee and hip, cyphosis, lordosis, and colossal scoliosis may develop as manifold consequences of atrophic paralysis. That subluxation of the humerus and the claw-hand may occur in the upper extremities has already been mentioned.

In Seeligmüller's 75 cases, 53, or 71 per cent., exhibited some kind of deformity. Among these, 43, or 56 per cent., were of the foot; 6 were cases of subluxated humeri; 5, easily-reducible luxation of the fingers.

The following table contains a summary of the deformities observed as a consequence of atrophic paralysis. They are distinguishable from congenital deformities dependent on altered relations of articular surfaces through defective development⁶¹ by being easily reducible. This remark especially applies to paralytic club-foot:

⁶¹ Volkmann's Handbuch, Billroth und Pitha.

Mechanism of Deformities.—From what has been said on the cases in which deformities are absent it is evident that one at least of three conditions are required for their production: the paralysis must be unequally distributed in the muscles surrounding a joint; pressure must be exerted by the weight of the body or traction by the weight of the limb; effort must be made to utilize the maimed part of the limb by means of other parts, or even by the muscles of the trunk.

The share taken by these different factors in the production of deformities has been differently estimated by different observers. The French surgeon Delpech was the first to explain the phenomenon on the theory of muscular antagonism. The same theory has been most minutely elaborated by Duchenne.⁶² According to it, the intact or less paralyzed muscles, in virtue of their tonus, constantly tend to draw the segment of the limb on which they act in a direction opposed to that in which it should be drawn by the paralyzed muscles. Since this action is unantagonized, its influence persists; the insertion-points of the contracting muscle being permanently approximated, the nutrition of the muscle is modified: it grows shorter (adapted atrophy). There results finally shortening and retraction of the muscles on one side of the joint, over-stretching of those on the other.

⁶² De l'Électrisation localisée, 1861.

Duchenne used to illustrate this theory by means of a skeleton supplied with artificial muscles, whose successive section would cause the appearance of the corresponding deformity. Werner⁶³ first protested against this theory, and the protest has been further developed and a different theory built up through the efforts of Hueter⁶⁴ and Volkmann.⁶⁵ The latter observes that the muscles and ligaments surrounding a joint normally receive a large amount of the weight falling upon its articular surfaces. Removal of this elastic resistance exposes these surfaces to the full force of the pressure, and thence to almost certain danger of deformity—a danger, therefore, always incurred after paralysis of the muscles. Thus, the weight of the body, pressing, unresisted, on the arch of the foot, is able to displace the bones of the arch from their normal relations and completely flatten the arch.

⁶³ Reform der Orthopædie, 1845.

⁶⁴ Gelenkkrankheiten.

⁶⁵ Sammlung klin. Vort., No. 1.

A position of ease is that in which the movement of the joint has been pushed as far as possible until limited by the passive resistance of the ligaments or the conformation of the articular surfaces. The weight of the body must then be so placed that the line of gravity falls on the side of the open angle, while the limiting bands stretch across the base. Thus, a tired man sits with a curved back; the muscles which may extend the spinal column in a straight line are relaxed; the column falls forward until arrested by the anterior vertebral ligaments. Thus, in standing at ease the thigh presses against the leg, so as to form a wide angle open anteriorly. When the quadriceps extensor is paralyzed, this position is inevitable and exaggerated, since the force which might counteract it, contraction of the thigh extensor, has been removed.

Formerly, the rôle of muscles in this elastic resistance was under-estimated and that of ligaments exaggerated.

CRITICISM OF THE THEORY OF MUSCULAR ANTAGONISM.—Three considerations have been urged in objection to the theory of muscular antagonism: First, deformities may develop even in limbs totally paralyzed, provided these limbs be subjected to weight and pressure. Pes equinus, the most common form of paralytic club-foot, develops with total paralysis of the muscles of the leg where the child does not walk, but is carried on the arms of a nurse with its foot dangling. The part of the foot anterior to the ankle-joint being longer and heavier than that behind, the point falls; the tendo Achillis is passively shortened, and by nutritive adaptation to this position may become permanently retracted. Long persistence in this position accustoms the dorsal surface of the bones to a less degree of pressure than the plantar surface: as a consequence, the growth of the bone becomes more active above, while it is arrested below; the arch of the foot is increased until the sole is curved into a deep hollow; and the plantar aponeurosis is correspondingly shortened.

Volkmann relates a case where this same deformity appeared without the least paralysis, but simply from prolonged passive extension of the feet in bed. The patient was an adult, and suffered from a severe typhoid with a double relapse. After recovering from the fever a year of orthopædic treatment was required to restore the feet to their normal position.

The second objection is the absence of any proof of such constant tonus in the muscles as may be sufficiently powerful to determine the position of a limb. Such tonus exists in the involuntary muscles, especially in those of the blood-vessels, but there is no evidence that it exists in the voluntary muscles. To this Seeligmüller has replied by admitting the objection to the theory as thus proposed, but substituting the more plausible influence of repeated contractions on the part of the non-paralyzed muscles. Each contraction draws the limb in a certain direction, and there it tends to remain, because there is nothing to antagonize the force which it has obeyed.

The third objection is that examination of individual cases not infrequently shows displacements in directions opposed to that which should be determined by muscular antagonism. Volkmann has especially illustrated the latter assertion by the mechanism of genu-recurvation. In paralysis of the quadriceps extensor of the tibia the weight of the body is exercised, not merely from above downward, but from without inward, falling, therefore, on the inner malleolus. It thus tends to press the anterior part of the foot outward,⁶⁶ and a valgus finally complicates the calcaneus. Seeligmüller, however, quotes two cases of pes calcaneus developed in children who had never walked: in one paralysis of the sural muscles had occurred at the age of four weeks, and the other case was observed at the age of fourteen weeks. Seeligmüller remarks that only early examination of the faradic contractility of a group of paralyzed muscles can decide whether any among them preponderate during a time sufficient to fix the limb in a vicious position. Thus in one case of pes calcaneus he found two years after the occurrence of the paralysis that some contractility still persisted in the dorsal flexors, but six months later this had quite disappeared. Had the examination then been made for the first time when all the muscles were equally paralyzed, it would have seemed impossible to explain the deformity by muscular antagonism.

⁶⁶ Sayre asserts that lateral rotation cannot take place at the ankle-joint, but at the medio-tarsal articulation. Hueter also refers pes varus and pes valgus to the talo-tarsal articulation.

RELATION OF WEIGHT AND MUSCULAR FORCES.—The influences of weight and of muscular action sometimes concur, sometimes are opposed to each other. Thus, the weight of the foot alone always tends to produce equino-varus; it acts therefore to intensify the action of the sural muscles when the anterior tibial are paralyzed, but to diminish the influence of paralysis of the gastrocnemius when the foot is being drawn into dorsal flexion. Hence one reason for the comparative rarity of pes calcaneus.

Paralytic deformities at the hip and knee are much rarer than those of the foot. At the hip this immunity is partly due to the relative rarity of paralysis in the muscles surrounding the joint—still more to the fact that the weight of the limb tends to correct excessive flexions. These are therefore more likely to occur in children allowed to remain in bed than in those who are encouraged to walk by means of suitable apparatus. The use of crutches, however, favors the development of deformity, because, since with paralysis of the thigh- or leg-muscles pes equinus nearly always exists, the thigh is unduly lengthened. To palliate this inconvenience the patient instinctively flexes the knee or hip, or both, and the position tends to become permanent.

When the flexion is rigid and extension becomes impossible, the gastrocnemii are relaxed until they lose their power of tension, and thence of fixing or raising the heel. Further, as by the flexion the limb is moved in front of the body, it is necessary to project the body forward again over the support. Hence a lordosis is developed, to be distinguished from that caused by paralysis of the vertebral extensors (a) by the rigid flexion of the thigh; (b) by the facility with which the patient can extend the back as soon as he is placed in a sitting position.

Although the quadriceps extensor is so frequently paralyzed, rigid flexion at the knee—such as on the theory of muscular antagonism might be expected from the action of the hamstring muscles—is very rare. As already observed, it occurs, if at all, in neglected children allowed to lie or sit with the leg partly flexed. In those who attempt to walk the leg is not flexed, but forced into hyper-extension by the following mechanism: The muscles inserted on the upper part of the thigh swing the leg forward like a passive support. Afterward the body bends forward over the support, and its weight, pressing from above downward and from before backward, and pressing the articular surfaces of the joint together, forces the head of the tibia backward until the movement is checked by the posterior ligaments. The deformity is the same whether the quadriceps or one or all the muscles surrounding a joint are paralyzed (Volkmann).

DISLOCATIONS.—Dangling limbs are, however, much more frequently the consequence of total paralysis, with extraordinary relaxation of the ligaments of the joint permitting dislocation. Reclus⁶⁷ has published several such cases. Verneuil has even suggested that congenital luxation of the hip-joint always depends on an intra-uterine spinal paralysis.⁶⁸<

⁶⁷ Revue mensuelle de Méd. et de Chir., Mars, 1878.

⁶⁸ Quoted from Seeligmüller.

In the upper extremities rigid contractions are much less frequent, even in proportion to the number of paralyses. Seeligmüller has seen five cases of reducible flexion of the fingers, and one of permanent extension of the wrist and fingers, associated with paralysis of all the flexors. In this case, if the arms were so suspended that the hand hung freely, its weight gradually overcame the action of the extensors and the fingers fell into flexion. Upon any attempt at exertion the hyperextension was reproduced.

DEFORMITIES OF THE TRUNK.—Scoliosis will be caused when, with unilateral paralysis of the extensors of the vertical column, the lower part of the trunk is drawn to the non-paralyzed side, and the upper half is bent over the paralyzed side in order to restore the balance. In bilateral paralysis of the extensors both shoulders are projected backward, so that a plumb-line dropped from between them falls behind the sacrum, and lordosis is developed, although the lumbar column is not projected forward. The same form of lordosis occurs when the glutæi are paralyzed. When, however, the abdominal muscles are paralyzed, the lumbar column is really projected forward, and then a plumb-line dropped from the shoulders passes over the sacrum.⁶⁹

⁶⁹ Duchenne, loc. cit., 1861.

PARALYSIS OF THE ABDOMINAL MUSCLES.—Unilateral paralysis of the extensors of the back is often difficult to detect in young children. Seeligmüller recommends that the child be laid across the mother's knees and told to move the trunk from one side to the other while the pelvis is held firm. If too young to obey the direction, the movement can be excited by pricking or by electric irritation. It will be seen that the trunk can be turned only to one side.

In these paralyses of the trunk it is clear that the deformity does not develop under the influence of muscular antagonism alone, but only when the non-paralyzed muscles attempt to sustain the superincumbent weight of the body or a portion of it.

SUMMARY OF MECHANISMS OF THE DEFORMITIES.—We may indeed conclude, as stated at the beginning, that this complex etiology exists in almost all cases. When the limb is at rest in bed or the weight of the body is transferred to crutches, then repeated contraction of the flexus will suffice to bring the limb into a vicious position (contraction at knee- and hip-joint). When the foot or hand hangs unsupported, its weight is sufficient to cause deformity, even when all the muscles are paralyzed, and sometimes in opposition to the direction of intact muscles. Thus the weight which is passively borne by the limb, and the efforts of intact muscles to effect the function of the limb in spite of the paralysis, both concur in the production of the deformity.

ANATOMICAL LESIONS.—The theory of the anatomical basis of infantile paralysis constitutes one of the most interesting portions of its history. It is indeed one of the most instructive chapters of modern pathology from the rapidity with which in a short time precise knowledge has accumulated, and for the degree in which this has revolutionized previous ideas.

Autopsies.—Until 1863 only five autopsies had been made upon persons affected with infantile paralysis.⁷⁰ As the disease was never, of itself, fatal, opportunity for pathological investigation did not present itself, while attention was still vividly attracted to the paralysis. When this had become a chronic infirmity, and patients had succumbed to intercurrent disease, the opportunity must have occurred, but was not then utilized.

⁷⁰ I believe the first list of modern autopsies was made by myself in a paper read before the New York County Medical Society, Dec. 22, 1873. The paper was not published until May, 1874, and in the mean time Seguin's lecture on “Infantile Paralysis” was published in the N. Y. Med. Record, Jan. 15, 1874, with a tabulated list of 25 autopsies, necessarily almost identical with mv own—cases 9 and 10 are really identical. A case by Roth is included; 4 negative autopsies are omitted. In 1880, Seeligmüller, in his elaborate essay in Gerhardt's Handbuch, published a list of 32 cases, including the above, and adding to them the following: Case by Müller, 1871; 4 cases by Leyden, 1875 (Arch. de Psych.); 1 case by Raymond, 1874 (Gaz. méd. de Paris); 1 case by Demme, 1876 (B. med. Bericht über das Jennerischt Kinderspital zu Bern); 1 case by Eisenlohr, 1876 (Tageblatt des Hamburger Naturforscher Versammlung); 1 case by Schultze, 1877 (Virch. Arch., Bd. lviii.).

To this list we may now add 3 cases by Turner, Humphrey, Taylor, 1881 (Path. Trans. London); 2 cases by Dejerine, 1878 (Progrès méd.), giving a total of 37 cases with detailed histories and autopsies.

The first cases affording autopsies in which the spinal cord was thoroughly investigated, belonged, however, precisely to this class of extremely chronic lesions, which should be regarded as showing rather the results of the morbid process than that process itself. Omission to observe this distinction has been the occasion of several misinterpretations of the pathological appearances.

Premature Theories.—In the entire absence of evidence it is a little remarkable that such high authorities as Rilliet and Barthez,⁷¹ West,⁷² Vogel,⁷³ Eulenburg,⁷⁴ should have pronounced dogmatically that the disease was essential—i.e. unaccompanied by any structural lesion whatever; and that Bouchut,⁷⁵ on the strength of most incomplete examination, should have built up a theory of myogenic paralysis. It is still more remarkable, after the published autopsies of Rinecker,⁷⁶ Laborde,⁷⁷ Cornil,⁷⁸ Prévost,⁷⁹ Charcot and Joffroy,⁸⁰ Parrot and Joffroy,⁸¹ Vulpian,⁸² Roger and Damaschino,⁸³ that Politzer in 1866,⁸⁴ Brown in 1871,⁸⁵ Barwell in 1872,⁸⁶ Kétli,⁸⁷ Adams in 1873,⁸⁸ should still adhere to this doctrine. They are, however, entirely in the minority, and all recent monographs and works published with a view to presenting the state of science assume the spinal nature of infantile paralysis to be established beyond possibility of controversy.⁸⁹

⁷¹ Traité des Maladies des Enfants.

⁷² Diseases of Children, Am. ed., 1860.

⁷³ Diseases of Children, trans. from 4th ed., 1870.

⁷⁴ Arch. Virchow, Bd. xvii. 1859.

⁷⁵ Deutsche Klinik, 1863.

⁷⁶ Gaz. méd., 1864, Soc. de Biol., 1864.

⁷⁷ De la Paralysie de l'Enfance.

⁷⁸ Gaz. méd., 1866, Soc. de Biol., 1866.

⁷⁹ Arch. de Phys., 1870, p. 134.

⁸⁰ Ibid., p. 310.

⁸¹ Ibid., p. 316.

⁸² Bouchut is said by Simon (loc. cit.) to be alone in his theory; Kétli, however, agrees with him. Eulenburg, in 1872, in his systematic treatise, assigns a central origin to infantile paralysis, but offers no opinion in regard to its nature.

⁸³ Gaz. méd., 1871.

⁸⁴ Jahrbuch für Kinderkrankheiten, 1866.

⁸⁵ Compend. für Kinderkrank., p. 161.

⁸⁶ Lancet, 1872.

⁸⁷ Jahrbuch für Kinderkrank., 1873.

⁸⁸ Treatise on Club-foot.

⁸⁹ See Seguin, loc. cit., 1874; Erb, Ziemssen's Handbuch, Bd. xi.; Seeligmüller, Gerhardt's Handbuch der Kinderkrankheiten; Ross, Treatise on Diseases of Nervous System, vol. ii.; Hammond, Diseases of Nervous System, 6th ed., 1881, etc. etc.

The following table contains a summary of the seven autopsies in which the spinal cord is said to have been examined with negative results. Of these, the only really important case is the third, in which a microscopic examination, made by so competent an histologist as Robin, was said to have discovered no lesion of the cord.

The foregoing autopsies may be tabulated as follows:

Heine, in the absence of autopsies, but arguing from clinical symptoms alone, already inferred the existence of a spinal lesion as cause of the paralysis, and believed that it consisted in congestion, or even in hemorrhagic exudation, capillary or massive, which should compress the cord and result in partial atrophy. The same opinion is advanced in 1844 by Brunnière,⁹⁰ also by Vogt,⁹¹ in 1868 by Salomon⁹² and Radcliffe.⁹³ The autopsies contained in the following table, in all of which vascular lesions are prominent, might be invoked in support of this view:

⁹⁰ Krankheiten des Gehirns und Ruckenmarkes.

⁹¹ Lahmung der Kinder, p. —.

⁹² Jahrb. f. Kinderheilk., 1865.

⁹³ Reynolds's Syst. of Medicine.

The first really modern autopsy, that made by Cornil in 1863, agrees with the two earliest on record in disclosing only an atrophy of the cord. Some of the other cases, contained in Table III., note in addition sclerosis of the lateral columns.

In all the remaining autopsies on record are noted atrophy of the anterior gray cornua and more or less extensive destruction of the ganglionic cells.

In a certain number of cases the atrophy seemed to be the unique lesion. But it never was confined to the ganglionic cells, but included the reticulum of gray fibres in which these were imbedded, and which was replaced by a reticulum of connective tissue. These cases were all examined many years after the occurrence of the paralysis.

The first group is contained in the following table:

In the remaining cases the ganglion-cells of the anterior cornua had also disappeared; but in addition to this atrophy excited distinct evidence of more or less extensive inflammation. This table includes one case of autopsy at two months (Roger's); one at six weeks after paralysis (Turner's).

It will be useful to add another table, which will group together the cases in which the autopsies were made within two years after the occurrence of the paralysis. Of these, all but the two made by Laborde, in which the cornua are declared to be healthy and the lesion limited to the white columns, show traces of destructive morbid processes in the gray substance of the cord, greatly predominating in the anterior cornua, but not absolutely limited to them, nor even to the part of the cord which corresponds to the paralyzed limb: