ON 
 
 TABES DORSALIS 
 
 TLbc Xumleian Xectures 
 
 DKLIVEKED BEFORE THE ROYAL COLLEGE OF PHVS1CL\NS, LONDON, 
 MARCH. 1906 
 
 DAVID FERRIER, M.D., LL.D., F.R.S. 
 
 /'ellow of the College ; Professor of Netiro-pathology in Kings College 
 
 Physician to Kings College Hospital, and to the National Hospital 
 
 for the Paralysed and Epileptic 
 
 ILonliou 
 JOHN BALE, SONS & DANIELSSON, Ltd. 
 
 OXFORD HOUSE 
 
 8391, (;t. titch field street, OXP'ORD STREEI", w. 
 
 1906
 
 \ -
 
 
 SIR RICHARD DOUGLAS POWELL. BART. 
 K.C.V.O.. M.D, . 
 
 PRESIDENT OF THE ROYAL COTLEGE OF PHYSICIANS, LONDON 
 
 THESE LECTURES 
 
 DELIVERED AT HIS REQUEST 
 
 ARE DEDICATED 
 
 AS A MARK OF ESTEEM
 
 CONTENTS. 
 
 Lecture I. 
 
 Introductory — -Historical Notes — Constitution of Posterior 
 Columns of Spinal Cord — Exogenous and Endogenous Tracts 
 — Morbid Anatomy of Tabes — Appearances in Spinal Cord at 
 different Levels — Changes in Spinal Ganglia, Peripheral 
 Nerves and Meninges — Theories of the Pathogeny of Tabes. 
 
 Lecture IL 
 
 Cerebro-Spinal Fluid in Tabes— Tabes a Dystrophy — 
 Relations to General Paralysis of the Insane — Etiology of 
 Tabes — Relation to Syphilis — Accessory Causes — Comparison 
 of Tabetic Degenerations with those caused by other Toxins 
 — Bacillary Origin of Tabes discussed. 
 
 Lecture III. 
 
 Physiological Pathology of Tabes — Ataxy — Its Sensory 
 Origin — Forms of Sensory Impairment in Tabes— Muscular 
 Hypotonia — Coordination of Movements — The Tabetic Pupil 
 — Theories as to its Pathogeny — Relation to Ciliary Ganglion 
 ■ — Conclusion— Bibliography.
 
 THE LUMLEIAX LECTURES 
 
 ON 
 
 TABES DORSALIS. 
 
 DELIVERED BEFORE THE ROYAL COLLEGE OF PHYSICLWS, 
 
 1906. 
 
 Lecture I. 
 
 Mr. President and Gentlemen. — The choice 
 of a subject must always, I imagine, be a matter 
 of difficulty to the one who has the honour of being- 
 appointed to deliver the Lumleian Lectures. Hav- 
 ing been invited to open a discussion on tabes at 
 the forthcoming International Medical Congress, it 
 appeared to me that I might with advantage to 
 myself, and 1 hope not without profit to you, take 
 as the basis of my lectures some questions relating 
 to the nature and pathology of this affection. 
 
 Tabes seems to be of perennial interest. Nor is 
 this much to be wondered at, considering the vast 
 held which it covers, and the variety of questions 
 which it raises. 
 
 The literature of tabes is enormous, and the 
 
 treatises and memoirs relating to it literallv count 
 
 by thousands. It will be my endeavour in these 
 
 lectures to presenc to you as succinctly as I can 
 
 I
 
 2 THE LUMLEIAN LECTURES 
 
 the evolution of our knowledge of tabes, its nature 
 and causes, and to indicate the problems as yet 
 unsolved and on which we need further light. 
 
 Tabes dorsalis, in the modern acceptation of the 
 term (and in reference to this I would remark at 
 the outset that but for their consecration by custom. 
 I should like to discard the term dorsalis altooether, 
 and the still more barbarous adjective tabetic, in 
 favour of tabes and tabid) is a very different thing 
 from the tabes dorsalis or dorsualis — ^Biai^ voiruU — 
 of the Hippocratic writers. This was essentially a 
 condition of neurasthenia associated with sperma- 
 torrhoea, and attributed to sexual vice or excess. 
 It was supposed to be due to wasting of the spinal 
 cord, but on what evidence this was based there is 
 nothing to show. That true tabes ever existed in 
 Europe before the introduction of syphilis in the 
 fifteenth century is more than doubtful. The reten- 
 tion of the name tabes dorsalis (or dorsualis) by the 
 older German writers for the disease as we now 
 understand it, though strictly correct according to 
 modern pathology, tended largely to create con- 
 fusion and excite prejudice against its unfortunate 
 victims. Though cases of what we now know must 
 have been tabes had been described by various ob- 
 servers, such as Ollivier (1824), Hutin (1827), and 
 Cruveilhier (1835-42), (who figures the post-nioi-teni 
 appearances in his atlas), in France; and by Horn 
 (1827), Jacoby (1842), Steinthal (1844), &c., in 
 Germany; and though in this country Todd (1847) 
 had drawn a clear distinction between paralysis and
 
 ON TABES DORSALIS 3 
 
 incoordination,* and was the first to connect inco- 
 ordination with disease of the posterior columns of 
 the spinal cord, yet there is no doubt that the first 
 systematic account of the etiology, symptomatology, 
 diagnosis, prognosis, and treatment of tabes was 
 given by Romberg (1840 to 1857). In \\\'s, Lehrbuch 
 der N erven- Krankheiteii he describes the charac- 
 teristic gait ; the pathognomonic symptom now 
 called by his name ; the increase of the ataxic 
 disorders on shutting the eyes ; the shooting pains, 
 aniaesthesiae and parasthesia.' ; the bladder troubles ; 
 the affections of vision, and the striking myosis and 
 fixity of the pupils. 
 
 He mentions the relative infrequency of tabes in 
 women ; and as to prognosis, he utters the gloomy 
 verdict : " To none affected by this malady is there 
 any hope of recovery ; ilber alle ist der Stab 
 gebrochen'' 
 
 Of the morbid anatomy, he says that, notwith- 
 standing considerable differences, the post-inorteiu 
 examination shows for the most part partial atrophy 
 of the spinal cord, mainly in the lower part of the 
 lumbar enlargement and onwards, and in the nerves 
 issuing from it. The diminution of volume amounts 
 
 * Todd {Cyclopadia, vol. iii., p. 621) says two kinds of 
 paralysis may be observed in the lower extremities — " the 
 one consisting simply in the impairment or loss of voluntary 
 motion, the other distinguished by a diminution or total loss 
 of the power of co-ordinating movements." In two cases of 
 the latter affection he predicted disease of the posterior 
 columns of the spinal cord, to which he attributed the faculty 
 of harmonising the movements of the limbs with each other.
 
 4 THE LUMLEIAN LECTURES 
 
 to the half or two-thirds of a normal cord, affects 
 the grey and white substance, or only one of these. 
 The atrophy of the nerves in the leash of the cauda 
 equina is often so pronounced that only the empty 
 neurilemma -sheaths remain. Also the roots of 
 higher inserted nerves participate in the atrophy ; 
 and what is of particular interest, the posterior 
 roots are sometimes alone affected, at other times 
 along with the posterior columns of the cord, while 
 the anterior roots appear normal. 
 
 Romberg, however, did not clearly differentiate 
 the basis of ataxy from that of muscular paresis or 
 paralysis, and he describes as equally belonging 
 to tabes leathery induration of the white substance 
 and, more frequently, softening of the grey matter. 
 
 There was no notable addition to the literature 
 of tabes until Duchenne in 1858 published his 
 researches on " ataxie locomotrice." These were 
 the starting-point of the extraordinary amount of 
 attention that the subject has attracted since his 
 time. Duchenne, purposely endeavouring to divest 
 himself of all preconceived notions, and pursuing 
 his investiofations in the limbo of so-called oreneral 
 spinal paralysis and unclassified forms of chronic 
 myelitis, discovered what to him was a new disease, 
 which, from its most striking symptom, he termed 
 "ataxie locomotrice," adding the further term "pro- 
 gressive " on account of its habitual tendency to 
 advance rather than recede. He clearly differen- 
 tiated the affection from muscular paralysis or 
 paresis, and described with a masterly hand its
 
 ON TABES DORSALIS 5 
 
 most characteristic manifestations. Trosseau, in 
 his brilliant clinical lectures, largely spread abroad 
 the knowledge of this affection, but he did scant 
 justice to previous writers in calling it Duchenne's 
 disease. For, though practically a discovery on 
 Duchenne's part, it was no discovery in the true 
 sense of the term, as the malady was undoubtedly 
 the same as had been known under the name of 
 tabes dorsalis. And though Duchenne differen- 
 tiated the affection more precisely than had been 
 done before him, yet he contributed nothing to the 
 pathological anatomy of the disease, and was in- 
 clined to attribute it to some affection of the cere- 
 bellum rather than that of the spinal cord. 
 
 Trousseau went so far as to look upon the malady 
 as a pure neurosis, and considered such changes as 
 had been observed by others in the spinal cord and 
 posterior roots as probably only the secondary result 
 of the functional disturbances created by some 
 obscure general condition. This view was main 
 tained by others of his school, notably by Isnard 
 who regarded tabes as a neurosis of the muscular 
 sense. 
 
 The connection between the symptoms of tabes 
 or locomotor ataxy and degeneration of the pos- 
 terior columns, first termed grey degeneration by 
 Cruveilhier, had, however, been satisfactorily estab- 
 lished by many observers, among others by Todd 
 and Gull in this country. 
 
 The minute anatomy of the morbid changes began 
 with the microscopical investigations of Virchow
 
 6 THE LUMLKIAN LECTURES 
 
 (1855) ^"<^ Rokitansky (185 7). Rokitansky des- 
 cribed it as a proliferation of the structureless, 
 ependyma-like connective tissue and stroma of the 
 nerve centres, which caused destruction of the 
 nerve elements proper, and resulted in induration 
 or sclerosis of the affected parts {schwielige 
 Degeneration). 
 
 The process was not inflammatory, however, as 
 there was no exudation ; and Rokitansky included 
 in the same term the end-results of various other 
 forms of lesion of the nerve centres. 
 
 But the first important microscopical investiga- 
 tions, specially in reference to the now well-defined 
 symptom-complex of locomotor ataxy, were made 
 by Bourdon and Luys {1861). 
 
 They defined the condition as sclerosis of the 
 posterior roots, the end-result of a chronic inflam- 
 matory process ; and similar views as to the inflam- 
 matory origin of the degeneration were held by 
 Charcot, Vulpian, Gull, and others, v. Leyden, 
 however, in his celebrated monograph, Die graice 
 Degeneration der hinteren Rilckeninarks - striiuge 
 (1863), combated the inflammatory origin of 'the 
 degeneration, and maintained that it had more 
 the characters of a simple atrophy, and that such 
 indications as there might be of chronic inflamma- 
 tion in the membranes and vessels were only 
 secondary or accidental. He regarded the process 
 as a peculiar degeneration of the sensory tracts of 
 the spinal cord, ascending generally from below 
 upwards.
 
 ON TABPZS DORSALIS 7 
 
 At this time, however, he did not, as he has done 
 in later times, causally connect the degeneration of 
 the posterior columns of the cord with that of 
 the posterior roots or sensory nerves, though he 
 specially noted their almost constant coincident 
 affection. In fact, in commentino- on a case re- 
 ported by Gull, to whom it seemed as if the 
 degeneration of the posterior roots caused by 
 inflammation of the pia mater had propagated 
 itself to the posterior columns, he says explicitly : 
 " But the view that the atrophy of the posterior 
 roots caused by inflammation of the pia mater leads 
 to a similar degeneration of the posterior columns 
 is very unlikely. Moreover, the disease of the 
 posterior columns progresses more from the pos- 
 terior periphery and middle line, rather than follows 
 the distribution of the roots." 
 
 The next great advance in the pathology of tabes 
 was made by Pierret (1871), who in the course of 
 his investigations observed that the primary lesions 
 of tabes consisted, not in a degeneration of the 
 posterior columns as a whole, but "in two sym- 
 metrical islets of small extent, situated in a special 
 region of the posterior columns." 
 
 " This region," he says, " which I propose to call 
 Ics rubans externes of the posterior columns, 
 appears on transverse section in the form of two 
 narrow bands skirting the internal aspect of the 
 posterior horns. It appears to contain two orders 
 of fibres — commissural and internal radicular. The 
 sclerosed areas do not actually touch the posterior
 
 ■8 
 
 THE LUMLEIAN LECTURES 
 
 cornua, but are separated from them by tracts of 
 healthy white substance, though they are connected 
 by fibrous bands. The sclerosis, beginning in the 
 middle of the rtiban externe, tends to spread 
 inwards towards the median column, and outwards 
 towards the posterior cornu " (see fig. r). 
 
 Ji!^4. 
 
 
 Fig. I. — Transverse sections of spinal cord in lower dorsal and 
 cer\'ical regions from a case of tabes (Pierret, Archives de Pliysio- 
 logic^ 1871). In lower dorsal region, note sclerosis in bandclettcs 
 externes^ also a small area of degenerations in GoU's columns. In cer- 
 vical region the bandelettes are much sclerosed ; Coil's columns much 
 less so. 
 
 Tabes then came for a time to be considered 
 a system disease or sclerosis of the nibans or 
 bandelettes externes of the columns of Burdach. 
 But at this epoch nothing was known as to the 
 relation of the bandelettes externes to the posterior 
 roots. V^ulpian, however {1879), saw the difficulty 
 of connecting the supposed primary disease of the 
 bundelettes exte^^nes with the atrophy of the pos- 
 terior roots. A primary degeneration of the 
 posterior columns could scarcely, he maintained, 
 lead to atrophy of the posterior roots whose trophic 
 centre was in the spinal ganglia. This would be 
 a retrograde degeneration and opposed to the
 
 ON TABES DORSALIS 
 
 VVallerian law. He therefore regarded the atrophy 
 of the roots as undoubtedly of primary orig'in. 
 
 Fig. 3. 
 
 Fig. 2. 
 Ascending degeneration of a lumbar root 
 in man, on right side. (Collier & Buzzard.) 
 
 
 C.ll. 
 
 
 C.VUl 
 
 D.rv 
 
 D.vn. 
 
 L.m 
 
 Fig. 3. 
 
 Ascending degeneration following section 
 of the posterior roots in the monkey from 
 L. 3 to D. 12 on left side. (Tooth, Gulsto- 
 nian Lectures, 1889.) 
 
 That of the bandelettes externes might perhaps 
 be so also, but only in so far as they contained
 
 lo thp: lumleian lectures 
 
 fibres not directly continuous with the posterior 
 roots themselves. 
 
 He proposed to substitute for the formula 
 "sclerosis of the posterior columns," "sclerosis of 
 the posterior roots," as the primary basis of tabes. 
 He thus inaugurated the radicular or root theory 
 of tabes, which is now entertained in one form or 
 another by the majority of neuro-pathologists. 
 
 In order to follow the bearing of this theory on 
 the pathological anatomy of tabes, it is necessary 
 to make some preliminary remarks on the relation 
 of the posterior roots to the posterior columns ot 
 the spinal cord. 
 
 The facts on which this is based have been 
 established mainly on the results of degeneration 
 consequent on lesions of the posterior roots by 
 disease in man (see fig. 2), and of experimental 
 section of the posterior roots in animals, especially 
 in monkeys. These we owe in part to the labours 
 of Schiefferdecker, Schultze, Kahler and Pick. 
 Singer and Munzer, &c. ; and notably to the careful 
 experiments and investigations of Tooth, which he 
 made the subject of his Gulstonian Lectures before 
 this College in 1889 (see fig. 3). 
 
 The posterior roots emerging from their respec- 
 tive ganglia (see fig. 4) penetrate the cord in two 
 divisions — an external, consisting of fibres of small 
 calibre which enter the apex of the posterior cornu, 
 or zone of Lissauer, and distribute themselves here ; 
 and secondly, an internal division, formed of fibres 
 of large calibre which alone take part in the forma-
 
 ON TABES DORSALIS 
 
 I I 
 
 Nucleus Cuneatus -■■ 
 Nucleus Gracilis — 
 
 Long Exogenous Ascending Fibre 
 (to Medulla) 
 
 Clarke's Cohmm 
 
 Medium Exogenous Ascending Fibre 
 (to Clarke's Column) 
 
 Keflex Collateral Fibre (to Anterior Cornii) 
 Short Fibre (to Posterior Cornuj 
 
 Fine Exogenous Fibrt 
 (to Lissauer's Tract I 
 
 Posterior Root Ganglion 
 
 ENDOGENOUS TRACTS 
 Descending %%S5 
 
 Ascending '^S^. 
 
 MEDULLA 
 
 MID-CERVICAU 
 
 UPPER DORSAL 
 
 MID-DORSAL 
 
 UPPER LUMBAR 
 
 LOWER LUMBAR 
 
 SACRAL 
 
 COCCYGEAL 
 
 Areas of Tabetic Degeneration 
 indicated by dots thus •^;- 
 
 Fig. 4.
 
 i2 THE LUMLEIAN LECTURES 
 
 tion of the posterior column. This division of the 
 posterior root gives off short, medium, and long 
 fibres. It occupies in the first place a crescentic 
 area on the inner border of the posterior cornu — 
 the cornu-radicitlar zone (Marie) — and gives off 
 short fibres to the adjoining cells of the posterior 
 cornu. With the advent of the next and successive 
 roots above, the lower roots are displaced from the 
 cornu-radicular zone inwards towards the middle 
 of Burdach's column, and occupy a space elongated 
 antero-posteriorly, which corresponds to what was 
 termed by Pierret the baiidelette externe. The 
 haiidelette externe is only the second stage in the 
 course of the posterior roots. From this are given 
 off the medium fibres, which form two principal 
 groups : first, the sensitivo-refiex group or refiex 
 collaterals of Kolliker, which arise mainly from 
 the middle of the bandelette, and are distributed 
 among the cells of the anterior cornu ; secondly, 
 the fibres to Clarke's column. These emerge 
 principally from the anterior aspect, and if derived 
 from the lowermost roots, ascend a considerable 
 distance before they finally reach their destina- 
 tion, inasmuch as the cells of Clarke's column 
 do not descend below the second lumbar segment. 
 Besides the short and medium fibres above des- 
 cribed there is another set of fibres which, instead 
 of ending in the cord, ascend in the column of 
 Goll as far as the nucleus gracilis of the medulla 
 oblongata. These fibres appear to come from the 
 posterior aspect of the baiidelette externe and
 
 ON TABES DORSALIS I3 
 
 postero-lateral zone, and oradually pass inwards, 
 and ultimately reach the mesial septum. The 
 lowest fibres are those which are nearest the 
 middle line. 
 
 Only the long fibres from the lumbo-sacral and 
 lower dorsal roots ascend in the column of Goll. 
 Those from the upper dorsal and the cervical roots 
 ascend in the column of Burdach, the lowermost 
 being" nearest the column of Goll, and end in the 
 nucleus cuneatus. 
 
 The columns of Goll. therefore, are formed by 
 the long fibres of the lumbo-sacral and lower dorsal 
 roots, and belong to the root system of the cord — 
 the exogenous or extrinsic system (Marie). All 
 these degenerate upwards on transverse section of 
 the spinal cord at any part. 
 
 But, in addition to the exogenous, or root fibres^ 
 the posterior columns of the cord contain tracts 
 which are probabl)- commissural in function be- 
 tween different segments and different levels ot the 
 spinal grey matter. These tracts constitute the 
 endogenous system of the spinal cord (Marie), and 
 form two divisions, the ascendino- and descending-. 
 They only degenerate on destruction of the grey 
 matter, and remain intact after all the exogenous 
 fibres consequent on destruction of the posterior 
 roots have undergone degeneration. 
 
 The ascending fibres occupy the cornn coniniis- 
 sural zone, or ventral field of the posterior columns, 
 lying posterior to the grey commissure, and extend- 
 ing sonie distance along the inner aspect of the
 
 14 THE LUMLEIAN LECTURES 
 
 posterior cornua. These vary in size according 
 to that of the horns of the lumbar, dorsal and 
 cervical regions, and degenerate above lesions of 
 the grey matter. 
 
 The descending system of fibres degenerates, 
 on the contrary, beloiv a transverse lesion of the 
 grey matter. This system of fibres w^as first 
 described by Schultze (1883), and is known in 
 the cervical and dorsal regions as the " comma " 
 tract. It occupies a position in the column of 
 Burdach, running obliquely in an antero-posterior 
 direction parallel to, but at some distance from, the 
 internal border of the posterior cornu. Schultze 
 described it as not reaching either the grey com- 
 missure or the posterior surface of the cord, but 
 the Marchi method has shown that this is not 
 quite correct. 
 
 This comma tract was supposed by him to con- 
 sist only of descending root fibres, but the researches 
 of various pathologists, especially those of Gom- 
 bault and Philippe, have shown that this comma 
 tract is continuous from above downwards with 
 other tracts which have been differently named 
 according to their appearance in different parts of 
 the cord — namely, the seplo-marginal tract of Bruce 
 and Muir in the lower dorsal region, the oval field 
 of Flechsig on each side of the mesial septum in 
 the lumbar region, and the median triangle of 
 Gombault and Philippe in the sacral region. 
 
 These variously-named tracts thus constitute one 
 continuous system of descending commissural fibres.
 
 ON TABES DORSALIS 
 
 15
 
 l6 THE LUMLEIAN LECTURES 
 
 receiving and giving off fibres at different levels 
 of the grey matter, and connecting together the 
 various segments, as well as the upper and lower 
 spinal centres, with each other. The continuity of 
 these systems with each other has been clearly 
 demonstrated by Purves Stewart in a case of crush 
 of the cervical cord. (See figs. 5 and 6.) 
 
 We may now proceed to consider the degenera- 
 tions characteristic of tabes, and to inquire how 
 far they correspond with the exogenous and endo- 
 genous systems of fibres above described. 
 
 In a moderately-advanced case of tabes one 
 observes a notable atrophy of the posterior roots. 
 These are thin, flattened and greyish in colour, 
 contrasting vividly with the white and plump 
 anterior roots. This appearance is most marked 
 in the roots forming the cauda equina. The 
 posterior columns of the cord are also thinned, 
 and of a greyish, semi-transparerit aspect, and 
 the posterior cornua are often similarly affected. 
 The pia mater, over the posterior columns more 
 especially, is frequently thickened and adherent. 
 
 The changes are, as a rule, most pronounced in 
 the lumbar and lower dorsal region, but may in 
 certain cases — so-called cervical tabes — be more or 
 less limited to the cervical region. In advanced 
 cases the whole of the posterior roots and whole 
 of the posterior columns may be affected. 
 
 If one examines a transverse section of the lower 
 dorsal or upper lumbar cord in an early case 
 (see fig. 7), one observes an area of degeneration,
 
 ON TABES DORSALIS 1 7 
 
 occupying- the middle third of the posterior column, 
 extending from the posterior cornu forwards and 
 inwards towards the posterior commissure. This 
 area constitutes Pierret's bandclette externe, and 
 varies in breadth according to the number of 
 posterior roots involved. The two bandelettes, not 
 always symmetrical, may meet almost in the middle 
 line. They correspond to the middle stage of the 
 traject of the posterior roots. From this part of the 
 
 *js6¥^-* '-^^ii-ta/*; 
 
 Fig. 7. — Transverse section in upper lumbar region from a case 
 of early tabes, showing degeneration in handelettes exfernes (Schmaus 
 and Sacki). 
 
 course, as we have seen, are given off the reflex 
 collaterals to the anterior cornua, and the fibres 
 CO the vesicular columns, or columns of Clarke. 
 These are the fibres first and specially affected in 
 tabes. 
 
 For a considerable time after the degeneration 
 has appeared in the bandelette externe, compara- 
 tively little degeneration may be visible in the 
 cornu-radicular or root-entry zone, in the long 
 fibres occupying the postero-external zones, and 
 
 2
 
 1 8 THE LUMLEIAN LECTURES 
 
 Still less in the cornu-commissural zone and oval 
 fields borderino- the posterior medium septum. In 
 more advanced cases, all the posterior columns are 
 degenerated except the cornu-commissural zone and 
 the oval fields of Flechsig- (see fig-. 8). These form, 
 as already remarked, the endogenous systems of the 
 cord, and only degenerate in very advanced cases 
 in which the upper regions of the cord are 
 invaded. 
 
 /'y 
 
 
 f<; 
 
 *^f ^^'^'""'•^ ^ 
 
 Fig. 8. — Transverse section of cord in an advanced case of tabes, 
 showing degeneration of the whole of the posterior columns, with the 
 exception of the endogenous tracts (Schmaus and Sacki). 
 
 The appearance of sections of the cord higher 
 up will vary according as the degeneration of the 
 posterior roots is confined to the lumbo-sacral 
 region, or also implicates the dorsal. 
 
 In the first case the only degeneration visible is 
 in the posterior median columns, which contain the 
 long fibres from the lumbo-sacral region in their 
 course up to the nuclei gracilis of the medulla 
 oblongata, all the short and median fibres having by 
 this time found their primary termination in the 
 cells of the posterior and anterior cornua and in
 
 ON TABES DORSALIS 
 
 •9 
 
 the columns of Clarke. The appearance presented 
 will be seen in fig. 9. 
 
 -. ?SVip( 
 
 Fig. 9. --Third cervical segment of cord from a case of tabes 
 (Leyden and Goldscheider, after Marinesco;. A, degenerated fibres 
 from sacro-lumbar roots ; B, degenerated fibres from dorsal and 
 lower cervical roots. 
 
 Fig. 10. — Fifth cervical segment from the same case. A, degenera- 
 tion of middle root zone ; B, degeneration of fibres to posterior horn 
 (collaterals) ; C, degeneration of Goirs column (kmibo-sacral) ; D, 
 degeneration of posterior root-zone fibres, newly entered. 
 
 If, on the other hand, the dorsal roots are also 
 affected, we get, in addition, degeneration in the 
 columns of Burdach, varying in extent according
 
 20 
 
 THE LUMLEIAN LECTURES 
 
 to the number of roots involved ; and we may 
 have a degenerated zone in Burdach's column 
 separated from that of the column of Goll by 
 
 Fig. II.— Section of mid-dorsal region. 
 
 Fig. 12. — Section of mid-lumbar region. A, atrophy of Clark's 
 column ; R, atrophy of posterior horn fibres ; C, middle root zone ; 
 D, ascending degeneration. (From same case as Figs. 9 and 10). 
 
 healthy tracts corresponding to healthy roots enter- 
 ing at a lower level (see fig. 13).
 
 ON TABES DORSALIS 
 
 21 
 
 In the cervical cord, if only the lower dorsal 
 and lumbo-sacral roots are affected, the degenera- 
 tion will be confined to Goll's columns, as these 
 are composed exclusively of the long fibres passing 
 up from these regions. If the upper dorsal is 
 affected, degeneration of Burdach's column will be 
 found immediately adjoining that of the column 
 of Goll ; and if the cervical roots are also im- 
 plicated, the degeneration will occupy the columns 
 of Burdach, varying in extent with the amount 
 
 Fig. 13. — Section of upper dorsal region showing tabetic degenera- 
 tion in the columns of Goll, separated by healthy areas from degenera- 
 tions in Burdach's columns (Schmaus and Sacki). 
 
 of degeneration of the roots. In cases of cervical 
 tabes there is no degeneration in the columns of 
 Goll (figs. 12 and 13). The degeneration is confined 
 exclusively to the columns of Burdach, and the long 
 fibres of these roots end in the nuclei cuneati. 
 
 It is only in the most advanced stages of tabes 
 that all the posterior columns, exogenous as well 
 as endogenous tracts, are alike degenerated. Of
 
 22 
 
 THE LUMLEIAN LECTURES 
 
 the exogenous system we have seen that they are 
 not all equally aftected at the same time. The 
 first incidence of the degenerativ^e process is on 
 the medium fibres, namely, the reflex collaterals and 
 
 Fig. 14. — Cervical tabes — third cervical segment (Leyden and 
 Goldscheider). 
 
 Fig. 15. — Cervical tabes — fifth cervical segment (Leyden and 
 Goldscheider). 
 
 those to Clarke's columns (fig-. 16). The long fibres 
 which, in large measure at least, constitute the 
 postero-external root zones are more resistant, and
 
 ON TABES DORSALIS 2^ 
 
 often exhibit scarcely any degeneration when the 
 bandelettes externes are completely sclerosed. 
 There is some doubt, however, whether the 
 postero-external root zones are composed only of 
 long fibres. There are grounds for believing that 
 they contain also endogenous fibres which are 
 
 Fig. 1 6. — Degeneration of posterior root fibres passing" towards 
 Clarke's column (Orr and Rows, Bra/;/, I904\ 
 
 relatively more resistant to degenerative change 
 (Schmaus and Sacki). 
 
 Similarly, the fine fibres of the posterior root 
 which enter the zone of Lissauer often escape 
 degeneration when it is already well marked in 
 the medium fibres of the column of Burdach (fig. 17) 
 
 The tabetic degeneration would therefore appear
 
 24 THE LUMLEIAN LECTURES 
 
 to be selective in character, unlike the Wallerian, 
 which affects all indiscriminately, attacking in suc- 
 cession fibres and tracts of perhaps different func- 
 tional significance. In this aspect it has been 
 attempted to establish a relation between the 
 tabetic degeneration and the foetal differentiation of 
 the various tracts which enter into the composition 
 
 Fig. 17. — Degeneration of posterior root-fibres passing into 
 posterior cornu. Lissauer's tract unafifacted (Orr and Rows). 
 
 of the posterior columns, as determined by the stage 
 of development at which they respectively receive 
 their myelin sheaths. Flechsig, by the myelinisa- 
 tion method, has divided the posterior columns 
 into six main divisions, which receive their myelin 
 sheaths at four successive stages of fcetal develop- 
 ment. He thus divides the posterior columns into
 
 ON TABES DORSALIS 25 
 
 an anterior, middle and posterior root zone ; this, 
 again, into a medial and a lateral. Added to these 
 are Goll's column, and a mesial tract bounding the 
 posterior median fissure. The first tract of fibres 
 to myelinate — his anterior root-zone — corresponds 
 with the cornu-commissural zone of Marie ; the 
 last to myelinate — his postero-lateral zone — corre- 
 sponds with the zone of Lissauer. These various 
 tracts constitute, according to Flechsig, elementary 
 systems, and it has been assumed that they are 
 composed of fibres of common functional signifi- 
 cance. The researches of Trepinski, however, on 
 the same subject, differ in many important respects 
 from those of Flechsig ; and there are many 
 discrepancies between both and the results of 
 secondary degeneration. Nor has it as yet been 
 satisfactorily shown that the tabetic degeneration 
 coincides precisely with any of the foetal systems 
 described by these authors. Further research, how- 
 ever, may clear up many doubtful points, for it 
 is not improbable that the tabetic degeneration 
 follows the lines of functionally different fibres, 
 which myelinate at different stages of foetal de- 
 velopment. 
 
 The tabetic defeneration, so evident in the 
 posterior columns and posterior roots, appears to 
 stop short at the spinal ganglia. The sensory 
 nerves distal to the ganglia have a normal appear- 
 ance and normal structure. But not infrequently, 
 as has been shown by Westphal, Pierret, Dejerine 
 and others, the finer peripheral ramifications in the
 
 26 THE LUMLEIAN LECTURES 
 
 skin, muscle, &.C., undergo atrophy similar to that 
 in the intramedullary terminations. Nor is the 
 process limited to the spinal cord and the spinal 
 nerves. The optic nerve is frequently affected, and 
 occasionally, also, other cranial nerves. Degenera- 
 tive changes have also been demonstrated in the 
 ciliary ganglia and in the sympathetic nervous 
 system, particularly in the cardiac plexus. These 
 will be considered more in detail in connection 
 with the physiological pathology of some of the 
 more prominent symptoms of tabes. 
 
 If we now turn from the more gross anatomy of 
 tabes and consider the intimate nature of the 
 process, we find, according to the most modern 
 research, that its essential character is a progres- 
 sive dystrophy, or demyelinisation and ultimate 
 destruction of the nerve fibres. The myelin sheath, 
 as shown by the Marchi method, breaks up, be- 
 comes granular, and finally disappears (figs. i6 
 and 17). The appearances vary somewhat accord- 
 ing to the acuteness of the process, but in general 
 a proliferation of the neuroglia network occurs 
 only in a proportion to the atrophy of the nerve 
 tissue proper. There is no active hyperplasia of 
 the neuroglia, and no vascular changes such as 
 can be considered as the primary starting point 
 of a sclerosis leading secondarily to atrophy of the 
 nerve fibres. The Marchi method in the early 
 stages of the process reveals the degeneration of 
 the myelin long before proliferation of the neu- 
 roglia has occurred, and such sclerosis as ulti-
 
 ON TABES DORSALIS 2 7 
 
 mately ensues is only what may be seen in any 
 organ consecutive to deoeneration of its true 
 parenchyma. 
 
 The obvious relationship of the exogenous tracts 
 of the spinal cord to the spinal ganglia — being, as 
 it were, the central prolongations to the sensory 
 neurons of which they are composed — naturally sug- 
 gests that the cause of the progressive dystrophy 
 of the posterior roots and their intramedullary 
 terminations may be found in some morbid altera- 
 tion of the cells of the spinal ganglia, the trophic 
 centres of the fibres in question. For, as Marie 
 remarks, it is difficult to believe that the branches 
 of a neuron, whether central or peripheral, and 
 performing the functions only of conductors, can 
 undergo degeneration apart from that of the cell 
 body which is their trophic centre. Therefore, in 
 every case of degeneration of a nerve fibre, whether 
 central or peripheral, the point first to investigate 
 is the celliLle inalade. 
 
 Much research has been spent on this point, but 
 the conclusions reached as to morbid changes in 
 the spinal ganglia, or their constituent cells, have 
 been far from harmonious. Gross changes, such 
 as were described by Bourdon and Luys, are uni- 
 formly absent. 
 
 As regards the cells, Marie himself found in 
 one case that they appeared less numerous than 
 normal ; that some of them seemed to show signs 
 of degeneration, and that there was distinct atrophy 
 of a considerable number of the nerve fibres
 
 28 
 
 THE LU.MLEIAN LECTURES 
 
 traversing the ganglia. Some support has been 
 given to Marie's view by the investigations of 
 Wollenberg, Oppenheim and Siemerling, Stroebe, 
 
 Fig. 19 
 Figs. 18 and 19. — Sections of Spinal Ganglia in advanced cases of 
 Tabes (Koster). 
 
 Redlich, Marinesco, Thomas and Hauser, &c. ; 
 but on the other hand, Schafferjuliusberger, Meyer, 
 Mott and many others have found them practically 
 normal.
 
 ON TABES DORSALIS 29 
 
 The alterations which have been described by 
 Wollenbergr, Marinesco, and others who have 
 obtained positive results, consist mainly in chroma- 
 tolysis, pigmentation or fatty changes in the cells, 
 occasional deformity or diminution in numbers, and 
 increase of the connective-tissue elements. Koster, 
 the most recent investigator on this subject, states 
 that in the early stages of tabes he has observed 
 appearances similar to those described by Wollen- 
 berg ; but that in more advanced cases there are 
 marked changes, such as chromatolysis, pigmenta- 
 tion, abnormalities in the nucleus, and distortion 
 of the cells (figs. 18 and 19). They were similar 
 to those resulting in course of time from experi- 
 mental section of the posterior roots in animals 
 (figs. 20 and 21). 
 
 It is therefore doubtful whether such changes as 
 have been discovered in the spinal ganglia afford 
 a sufficient anatomical substratum for the amount 
 of degeneration, more particularly in the posterior 
 roots and in their intramedullary terminals. And 
 it is considered (Schmaus) a powerful argument 
 against the spinal ganglion being regarded as the 
 primary origin of the dystrophy, that the peripheral 
 process, as it emerges from the distal pole of the 
 ganglion, practically shows no abnormality (with 
 the exception, perhaps, of some of its cutaneous 
 terminals) ; while the central process, the posterior 
 root, may be degenerated right into the ganglion 
 itself. Whether these and other arguments are 
 altogether valid against the primary ganglionic
 
 qo 
 
 THE LUMLEIAN LECTURES 
 
 origin of the tabetic process will be considered 
 later, but it must be admitted that the anatomical 
 changes hitherto actually demonstrated in the 
 
 Fig. 20. — Section of normal Spinal GanL^lion in Cat (Koster). 
 
 Fig. 21. — Section of Spinal Ganglion in Cat, one hundred and 
 twenty days after division of posterior root (Koster). 
 
 o-ano-lion cells are altoQether incommensurate with 
 the degree of degeneration of the processes (central
 
 ON TABES DORSALIS 3 I 
 
 and peripheral) as compared with that observable 
 in other forms of cell degeneration, such, for 
 instance, as in the multipolar cells of the anterior 
 cornua in atrophic spinal paralysis. 
 
 Instead of regarding the spinal ganglia as the 
 primary origin of the degeneration of the posterior 
 roots and exogenous medullary tracts, v. Leyden 
 and Goldscheider are inclined to look upon the peri- 
 pheral nerve endings, which not infrequently exhibit 
 dystrophic or neuritic changes, and are naturally 
 more exposed to traumatic influences, as more 
 probably the starting-point of the process. It is 
 known that after amputations and section of 
 nerves, atrophy occurs in the posterior horns, 
 posterior columns, and also in the anterior horns 
 of the spinal cord. This is the form of degenera- 
 tion termed retrograde, and is probably due to 
 the cessation of the normal influences which serve 
 to excite and maintain their functional activity. 
 And it has been shown by Lugaro, Cassirer, and 
 more recently by Warrington, that marked chroma- 
 tolytic changes occur in the cells of the spinal 
 ofano-lia on section of the sensorv nerves, and of 
 certain cells in particular, according to the kind 
 of afferent nerve — muscular or cutaneous — which 
 is divided. It is assumed, therefore, by the sup- 
 porters of the peripheral origin of tabes, that 
 changes, organic or functional, in the spinal gan- 
 glion cells may be induced by neuritic or other 
 processes in the peripheral nerve endings, and thus 
 secondarily lead to degeneration of the posterior
 
 J- 
 
 THE LUMLEIAN LECTURES 
 
 roots and posterior columns. It has not, however, 
 been demonstrated that degeneration of the peri- 
 pheral nerves always precedes that in the cord, 
 nor has it been shown that such peripheral de- 
 feneration is a constant feature of the morbid 
 anatomy of tabes, which it ought to be if the 
 theory is at all well founded. Nor does it seem 
 likely that neuritic or other degeneration, ascending 
 from the periphery, can induce such dynamical or 
 organic impairment of the ganglion cells as the 
 theory postulates, without making itself evident 
 at the point where the sensory nerves enter the 
 ganglia. The peripheral theory of the pathogeny 
 of the tabetic degeneration of the spinal cord has 
 therefore to overcome many serious objections 
 before it can be accepted as at all a satisfactory 
 one. 
 
 Other pathologists are of opinion that they have 
 discovered the origin of the tabetic process in 
 inflammatory conditions of the meninges affecting 
 the posterior roots. The theories of Nageotte, 
 and of Redlich and Obersteiner, are the most 
 important in this relation. 
 
 Nageotte, whose views have received a large 
 amount of support, especially in France, has 
 called attention to an anatomical disposition of 
 the posterior roots which renders them peculiarly 
 liable to noxious influences. Nageotte holds that 
 in tabes there is a diffuse chronic meningitis which 
 has all the characters of syphilitic origin — namely, 
 infiltration with lymphocytes and plasma cells.
 
 ON TABES DORSALIS 
 
 33 
 
 especially round the veins, a meningo-myelitic 
 affection of the periphery of the cord, medulla, 
 and even the cortex cerebri. 
 
 This general meningitis induces the degenera- 
 tion characteristic of tabes by implicating the 
 spinal roots at the point of their emergence from 
 the dural sac. The anterior and posterior roots 
 (fig. 22) after leaving the spinal cord traverse the 
 
 DM, 
 
 Fig. 22. — Diagram of Emergence of a Spinal Root from the 
 Dural Sac (Nageotte). Ar. = arachnoid ; D.M. = dura mater ; 
 Es.A. = sub-arachnoid space ; Esp. Lymph. = lymphatic space ; 
 G. = spinal ganglion ; N.Rad. = radicular nerve ; P.M. = pia mater; 
 R..\nt. = anterior root ; R. Post = posterior root; Ts.A. = arachnoid 
 tissue. 
 
 sub-arachnoid space, and at the point of exit from 
 the dural sac receive a sheath both from the dura 
 and arachnoid, which constitutes the peri-fascicular 
 sheath of the roots, and beyond the ganglion the 
 two roots unite to form the mixed nerve. The 
 portion which lies between the ganglion and the 
 dural sac, which varies in length in different 
 regions, he terms the radicular nerve. The peri- 
 fascicular sheath, derived from the dura and the 
 arachnoid, divides the radicles into compartments, 
 and the intra-fascicular tissue is continuous with
 
 34 THE LUMLEIAN LECTURES 
 
 the pia mater. The lymphatic spaces of the 
 radicles communicate with the subarachnoid space, 
 and constitute a narrow and complicated channel 
 between the great lymphatic sac and the efferent 
 vessels. Owing- to this disposition the radicles 
 are specially liable to injury by menino"eal inflamma- 
 tion and by toxins circulating in the cerebro- spinal 
 fluid. Hence is developed a transverse interstitial 
 neuritis, which specially affects the posterior radi- 
 cles, and the anterior also, to some extent, which, 
 however, is only transient. 
 
 In consequence of this neuritis, secondary de- 
 generation is set up in the intramedullary con- 
 tinuations of the posterior roots, and spreads back 
 to the spinal g-anglia. The process, however, is 
 not a true Wallerian degeneration, but a mere 
 dystrophy, or lowered vitality, which affects more 
 particularly the short and medium fibres of the 
 posterior roots. In support of his views, he 
 adduces the lymphocytosis of the cerebro-spinal 
 fluid which occurs in tabes, to which I will call 
 your attention presently. But in the meantime 
 Nageotte's theory appears to offer such a simple 
 explanation of the tabetic degeneration that it 
 seems almost a pity to subject it to destructive 
 criticism. 
 
 But it is open to the same objections as can 
 be urged against the theory of Redlich and Ober- 
 steiner, in that the meningitis, which it postulates, 
 is not a constant phenomenon, and is perhaps 
 more frequently absent than present, and when it
 
 ON TABES DORSALIS 35 
 
 exists is more often of the character of a secondary 
 thickening than an inflammatory process. And 
 further, it has been found in the early stages of 
 tabes that the intramedullary degeneration is often 
 well marked when the posterior roots exhibit no 
 appreciable change and no indications of a local 
 neuritis. The practical escape of the anterior 
 roots is, I think, a fatal objection to Nageotte's 
 
 Fig. 23. — Points of entry of the posterior roots into the spinal cord 
 (Redhch). 
 
 theory. That the neuritis which he also finds in 
 the anterior radicle should be evanescent, while 
 that of the posterior radicle leads to such pro- 
 found intramedullary degeneration, is unintelligible ; 
 and is, moreover, opposed to the usual relatively 
 greater affection of the motor than of the sensory 
 fibres when neuritis attacks mixed nerves. 
 
 Redlich and Obersteiner maintain that at the
 
 36 THE LUMLEIAN LECTURES 
 
 point where the posterior roots penetrate the pia 
 mater there is a constriction which constitutes a 
 loctis minoi'is resistciitiae (fig-. 23), and that in con- 
 sequence of meningeal inllammation and retraction 
 of cicatrisation, as well as by thickening of the 
 closely-adherent blood-vessels, the posterior roots 
 become, as it were, strang-led. Hence ensues 
 degeneration in the intramedullary exogenous tracts, 
 followed at a later stage by that of the posterior 
 roots in a retrograde manner. 
 
 Kedlich and Obersteiner's view has the merit 
 of harmonising better with the more common 
 starting-point of the tabetic degeneration than 
 that of Nageotte, but, as has been remarked, the 
 meningitis basis on which it rests is not very sub- 
 stantial. It is not universally present, nor is 
 meningitis, syphilitic or otherwise, when it does 
 occur, a common cause ot intraspinal degenera- 
 tion by implication of the posterior roots. That 
 such a result is possible, and perhaps actually 
 occurs in some cases, may be admitted, but it is 
 extremely rare. The symmetrical and segmental 
 character of the tabetic deoenerations, which it 
 would be difficult to correlate with any, so to 
 speak, brutal meningeal inflammation, and the 
 occurrence of degenerations elsewhere, such as in 
 the optic nerve, ciliary ganglion, and sympathetic 
 system, are all formidable arguments against the 
 meningitic origin of tabes, whether at the point 
 maintained by Nageotte or that of Redlich and 
 Obersteiner.
 
 ON TABES DORSALIS 
 
 .S7 
 
 But that the so-called "ring of Obersteiner' is 
 a locus viinoris resistentiae, or specially vulnerable 
 point, in the course of the posterior root fibres, 
 has been argued with much force by Orr and 
 Rows in their memoir on the system lesions of 
 the posterior columns in General Paralysis. 
 
 At this point the extramedullary fibres of the 
 posterior roots lose their neurilemma sheath, 
 
 Fig. 24.— Entrance of posterior root into the cord, showing ring of 
 Obersteiner ("white band crossing posterior root). (After Orr and 
 Rows.) 
 
 and the myeline sheath is also much thinned. 
 That the absence of the neurilemma sheath 
 renders the nerve more vulnerable is probable 
 from the researches of Kennedy, Bethe, Ballance 
 and Purves Stewart, which have demonstrated the 
 importance of the neurilemma and its cells in the
 
 ^S THE LUMLEIAN LECTURES 
 
 nutrition and regeneration of nerves. These 
 show that even in the distal segment of a divided 
 nerve regeneration of the axis-cyHnders and me- 
 dullary sheaths take place, although the full 
 maturity of the new nerve fibres is not attained 
 until union of the proximal and distal segments 
 occurs, and the fibres become functionally con- 
 tinuous with each other. This regeneration is 
 dependent on the cells of the neurilemma sheath. 
 Where these are absent, as in the intraspinal 
 fibres, regeneration does not occur ; and it is pro- 
 bable, therefore, that the absence of the neuri- 
 lemma cells is the cause of the greater proclivity 
 to degeneration of the intramedullary terminals 
 than of the other branches of the sensory proto- 
 neurone in the tabetic dystrophy. 
 
 It has been shown also that degeneration ot 
 the intramedullary root fibres, similar to those 
 of tabes, may occur in the course of cerebral 
 tumours, diabetes, aneemia, and other cachetic 
 states (Homen, Nageotte, Williamson. Bruce, 
 Erbsloh, &c.). The degeneration extends from 
 Obersteiner's ring; into the reflex collaterals and 
 fibres passing into Clarke's columns. Whether it 
 begins first in the terminal fibres, and gradually 
 extends backwards, or vice ve?'sd, is the subject of 
 some difference of opinion. Orr and Rows hold 
 that it commences at Obersteiner's ring and 
 extends inwards ; while Nageotte considers that 
 the terminal fibres of the reflex collaterals and 
 those of Clarke's columns are first affected. The
 
 ON TABES DORSALIS 39 
 
 more recent investigations of Spielmeyer, by means 
 of Cajal's new axis-cylinder stain, are in favour 
 of the view that the terminal fibres which arboresce 
 around the cells of Clarke's columns, as well as 
 those of the posterior column nuclei, first show 
 signs of degeneration. This is followed by glia 
 proliferation in proportion to the destruction of the 
 nerve filaments. 
 
 Whether we regard the intramedullary degene- 
 ration as centripetal or centrifugal, or more pro- 
 bably simultaneous, it is clear that this portion 
 of the posterior root is more liable to show signs 
 of impaired nutrition at an early stage than either 
 the extramedullary portion or the cells of the 
 spinal ganglia. Some conditions probably exist 
 favourino- the incidence of noxious influences on 
 this portion of the sensory protoneurone. In 
 this connection Marie and Guillain have laid 
 great stress on the peculiarities of the lymph 
 circulation in the posterior columns of the spinal 
 cord. They maintain that in the pia mater of 
 the posterior columns, between the outer and 
 inner layer, there exists a lymphatic system, the 
 spaces of which communicate with those of the 
 posterior columns, but are independent of those 
 of the antero-lateral columns. In this system the 
 current of lymph is upwards and towards the 
 central canal. In tabes they have found this 
 system dilated, and (by the Marchi method) black 
 granules in the cells of the ependyma in cases 
 where the central canal is permeable. They
 
 40 THE LUMLEIAN LECTURES 
 
 attribute the tabetic degeneration to a syphilitic 
 affection of the posterior lymphatics, and derive 
 the lymphocytosis of cerebrospinal fluid from the 
 elements of this system. 
 
 That noxious substances or toxins orig-inatinof 
 in the periphery ascend to the spinal cord more 
 readily by the posterior than by the anterior 
 roots has been shown by Homen, and also by 
 Orr and Rows ; and that in this way, or by 
 peculiarities in the lymphatic circulation in the 
 cord, the posterior roots and posterior columns 
 are specially exposed to injury is extremely pro- 
 bable. But that there is in tabes a posterior 
 meningitis of a syphilitic character, as maintained 
 by Marie and Guillain, is not generally admitted. 
 It is not supported by the results of anti-specific 
 treatment, nor can the lymphocytosis of the 
 cerebro-spinal fluid be adduced as a conclusive 
 argument in its favour.
 
 ON TABES DORSALIS 4 1 
 
 Lecture II. 
 
 In health the centrifugaHzed deposit from the 
 cerebro-spinal fluid contains scarcely any cellular 
 elements beyond an occasional endothelial cell and 
 perhaps one to three lymphocytes in a field, with a 
 magnification of 400 diameters (fig. 25). 
 
 Fig. 25. — Centrifuge deposit from normal cerebro-spinal fluid 
 (Purves Stewart). 
 
 In certain diseases the cellular elements are 
 present in excess. In acute microbic affections of 
 the cerebro-spinal meninges, as by staphylococcus, 
 pneumococcus, and diplococcus intracellularis, a 
 leucocytosis occurs, in which the leucocytes are of 
 the polynuclear type. On the other hand, in certain 
 chronic affections of the meninges, syphilitic, tuber- 
 culous, &c., an excess of leucocytes also occurs. 
 These are of the mononuclear type ; in Qther words 
 there is a lymphocytosis. It has been shown by 
 Widal, Sicard, Ravaut, and others — and their 
 observations have been confirmed by my own and
 
 42 THE LUMLEIAN LECTURES 
 
 Purves Stewart's investigations — that in tabes, 
 as well as in general paralysis of the insane, 
 from the earliest to the latest stacjes there is a 
 marked lymphocytosis of the cerebro-spinal tiuid. 
 I have in several instances relied on this lympho- 
 cytosis as establishing the diagnosis of general 
 paralysis before there were any abnormalities of the 
 pupils or other physical signs characteristic of the 
 disease. In a case of tabes or general paralysis, 
 instead of two or three, as in the normal state, one 
 may count a hundred, or even several hundred (fig. 
 26) lymphocytes in the field of the microscope. At 
 
 Fig. 26. — Centrifuge deposit from cerebro-spinal fluid in tabes 
 (Purves Stewart). 
 
 first sight this lymphocytosis would seem to confirm 
 the vnews of those who ascribe the disease to a 
 syphilitic affection of the meninges, either generally, 
 or of Nageotte's meningeal root sheath, or of the 
 posterior lymphatic system of the spinal cord, as 
 maintained by Marie and Guillain. But lympho- 
 cytosis occurs in other affections in which meningitis, 
 syphilitic or otherwise, plays no part. Thus it has 
 been found in Landry's paralysis, in the subacute
 
 ON TABES DORSALIS 43 
 
 combined degeneration of pernicious anaemia (Purves 
 Stewart), and in herpes zoster (Sicard). Lympho- 
 cytosis cannot, therefore, be regarded as pathogno- 
 monic of meningeal inflammation, though no doubt 
 meningeal inflammation would accentuate it. But 
 the most powerful argument against the syphilo- 
 meningeal origin of the lymphocytosis in tabes and 
 general paralysis, is the fact that it does not yield 
 in the sligfhtest decree to the most eneroetic anti- 
 syphilitic treatment (Milian, Purves Stewart). In 
 active syphilitic lesions of the nervous system, 
 however, the lymphocytosis that may be present 
 tends to disappear in the course of treatment. 
 Thus in a case of syphilitic paraplegia, under the 
 care of Purves Stewart, the lymphocytes after a 
 month's antispecific treatment had decreased from 
 52*6 to i6"i per field, and the symptoms had 
 become greatly ameliorated (figs. 27 and 28). 
 
 Having thus passed in review the different 
 theories which have been advanced in reference to 
 the pathogeny of the tabetic degeneration, we have 
 seen that there is not one which is not open to 
 several more or less serious objections. 
 
 Instead of attempting to be more precise than 
 the facts permit, I am inclined to adopt the 
 hypothesis of Thomas and Hauser, that the essential 
 lesion of tabes is a dystrophy, similar to that in- 
 duced by certain toxic agents, affecting the sensory 
 protoneurone as a whole, and manifesting itself in 
 degeneration of the peripheral and central termina- 
 tions, of which the intramedullary are the more
 
 44 THE LU.MLEIAN LECTURES 
 
 vulnerable, and are usually the earliest to exhibit 
 anatomical change. The process, however, is not 
 confined to the spinal protoneurone, but may affect 
 among- others the optic, the sympathetic, and certain 
 motor neurones. 
 
 Fig. 27. — Deposit from cerebro-spinal fluid in a case of syphilitic 
 paraplegia Purves Stewart). 
 
 Fig. 28. —The same case as fig. 27. after three weeks anti-syphilitic 
 treatment. 
 
 And here 1 would express, in concurrence with 
 Fournier, Mott. and many other neuro-pathologists 
 of the present day, my belief in the essential patho- 
 logical identity of tabes and general paralysis. 
 They are, in my opinion, merely different aspects 
 of the same polymorphic disease. One cannot say 
 thet thev are the same disease so far as their
 
 ON TABES DORSALIS 45 
 
 clinical manifestations and course are concerned, 
 any more than one would say that tuberculous 
 meningitis and pulmonary consumption are the 
 same. But the morbid process underlying both 
 is identical. 
 
 Both are tabetic or wasting affections — of the 
 sensory protoneurone in the one case, and of the 
 cortical neurones in the other. The essential 
 etiological factor is, as we shall see, the same, 
 and the date of development of the two affections 
 after the primary infection corresponds. No blind 
 or crude primary meningeal inflammation could 
 selectively attack only such regions — the fronto- 
 parietal by preference — as exhibit the atrophic 
 corte.x and adherent lepto-meninges characteristic 
 of general paralysis. 
 
 The primary lesion in general paralysis, as in 
 tabes, is a dystrophy of the neurones and their 
 connections ; the sclerosis, thickening of the 
 membranes and walls of the blood-vessels, being- 
 only a secondary result of the parenchymatous 
 degeneration. 
 
 Not infrequently tabes and general paralysis are 
 combined, or succeed each other, in the same 
 individual. Clinically, some of the characteristic 
 signs, notably the Argyll-Robertson pupil, are 
 common to both, and in both there is the same 
 cerebro-spinal lymphocytosis. 
 
 It would, in my opinion, conduce to a truer 
 conception of the pathology of these affections if 
 we were to call them both by the generic name
 
 46 THE LUMLEIAN LECTURES 
 
 "tabes," classifying- them specifically as Cerebral 
 Tabes (general paralysis), Spinal Tabes (tabes 
 dorsalis, locomotor ataxy), and Cerebro-spinal Tabes 
 (so-called tabo-paralysis), according as the disease 
 affects the brain, or spinal cord, or both conjointly. 
 
 Etiology of Tabes. 
 
 Up to 1876, when Fournier first pointed out the 
 great frequency of antecedent syphilitic infection in 
 tabes and indicated their probable causal relation- 
 ship, tabes was most commonly attributed to sexual 
 excess, or to the complex of injurious influences 
 associated with exposure to cold. The supposed 
 causal relationship between sexual excess and tabes 
 was probably only a daninosa hereditas from the 
 tabes dorsalis or sexual neurasthenia of the Hippo- 
 cratic writers. 
 
 Sexual excess, like all other debilitating influences, 
 may be placed among the exciting causes, and many 
 tabetics have probably exhausted themselves in 
 this way ; but the influence of sexual excess in the 
 causation of tabes has been greatly over-estimated, 
 and if that were the only cause, the disease would 
 certainly be much more common than it is. 
 
 Similarly as regards cold or damp cold, which 
 Romberg thought, and which is still regarded by 
 V. Leyden as one of the principle causes of tabes, 
 this view is probably founded more on the patient's 
 belief that his tabetic pains are rheumatic in 
 character, and that he must have caught a chill.
 
 ON TABES DORSALIS 47 
 
 slept in a damp bed, or the like, than on any more 
 substantial basis. When we consider how often, 
 when our etiology is at fault — not only in respect 
 of the diseases of the nervous system, but many 
 others — it has been the custom to fall back upon 
 exposure to cold as the probable cause, afterwards 
 proved insufficient, we may, without undue deprecia- 
 tion, call it the last refuge of destitute etiology. 
 Like sexual excess, exposure to cold and damp 
 would, without doubt, have an injurious influence, 
 and probably excite or intensify tabetic symptoms 
 already incipient, so that the patient might have 
 some grounds for ascribing his malady to this cause. 
 But that exposure to cold by itself is a sufficient 
 cause of tabes is, in my opinion, a mere vague 
 assertion, supported by no accurate data, and at 
 variance with ones daily clinical experience. A 
 history of exposure to cold is not found more 
 frequently in tabes than in many other affections of 
 the nervous system. 
 
 Fournier's views as to the causal relationship 
 between syphilis and tabes were at first received 
 with scepticism, not to say strong opposition, in 
 many quarters ; and the frequency of antecedent 
 syphilitic infection in the personal history of the 
 tabetic was not regarded as more than might be 
 found in other forms of disease of the nervous 
 system. But Fournier's views soon gained powerful 
 support, among the first by Gowers in this country 
 and by Erb in Germany ; and Erb, who at one 
 time (1878) was a strenuous opponent, is now
 
 48 THE LUMLEIAN LECTURES 
 
 perhaps the most eminent advocate of the syphiHtic 
 origin of tabes. It is now so generally admitted 
 by neuro-pathologists, with some few notable ex- 
 ceptions, such as V. Leyden, that syphilis is the 
 principle cause of tabes, that it seems unnecessary 
 to quote the data on which this opinion is based. 
 But as the elimination of cause and effect depends 
 for the present mainly on the statistical method, and 
 as this is liable to many fallacies depending on the 
 personal equation of the observers and the material 
 dealt with, it will be sufficient to state the results 
 obtained by Erb, whose probably unequalled ex- 
 perience and laborious inquiries, carried out with 
 systematic uniformity, invest his inductions with 
 exceptional value. 
 
 In the most recent of his numerous contribu- 
 tions on this question, Erb states that among 
 1,100 cases of tabes occurring in his own practice 
 among men of the better class, and therefore 
 better witnesses than those living among poor and 
 squalid surroundings, 89'45 per cent., or practi- 
 cally 90 per cent., had had chancre or syphilis ; 
 and only io"54 per cent., or practically 10 per cent., 
 had not been ostensibly infected. But of these, in 
 only 3 per cent, was it impossible to find any ground 
 for suspecting an antecedent syphilitic infection. 
 
 Now contrast these statistics, obtained from 
 1,100 tabetics, with those of 10,000 cases, also of 
 men of the better class, of all varieties of disease, 
 excluding tabes. Of these, only 21-5 per cent, had 
 had syphilis, as against 89'5 per cent, among the
 
 ON TABES DORSALIS 49 
 
 tabetics, that is, that syphilis was found 4*5 times 
 more frequently among tabetics than among- other 
 patients. Similar investigations among 138 tabetic 
 men of the lower orders give a somewhat lower 
 percentage of antecedent syphilis, namely, 77*2 per 
 cent. ; but among 1,300 non-tabetics only 6 '54 had 
 had syphilis, as against 93 "46 non-infected. 
 
 Among women, who are less liable to tabes than 
 men, and those of the better class less so than those 
 of the lower orders, it is, for various reasons, much 
 more difficult to determine the question of syphilitic 
 infection. But Erb found that amono i c tabetic 
 women of the better class, at least 867 per cent., 
 and probably 93*3 per cent., had been infected. 
 Among 16 women of the lower orders, however, 
 syphilis was probable in only 687 per cent., and 
 unproved in 3 i "3 per cent. 
 
 Taken altogether, a statistical inquiry among 
 1,321 cases — namely, 1,100 men of the better class, 
 gives 89*5 infected; among 158 men of the lower 
 class, 77*2 per cent. ; and among 63 women, 80 per 
 cent, as probably infected. Essentially the same is 
 true of general paralysis. 
 
 So far for mere statistics, the oeneral result beino- 
 that in about 10 per cent, of the cases syphilis is 
 unproved. These 10 per cent., however, constitute 
 the pivot round which the essential question turns, 
 whether tabes, like diphtheria or tuberculosis, is 
 always due to a specific infection, or whether it may 
 result from some other cause. This is not a mere 
 academic question, but one of far-reaching practical 
 
 4
 
 50 THE LUMLEIAN LECTURES 
 
 importance. Can we, in the present state of our 
 knowledge, give a definite, or approximately definite, 
 answer to it ? 
 
 We may, perhaps, do so by taking- other facts 
 into consideration besides mere percentages. 
 
 It has been established, and is generally recog- 
 nized, that tabes frequently follows syphilitic infection 
 of a mild type, and therefore more likely to escape 
 notice, particularly among those in whom personal 
 cleanliness is not a cardinal virtue. And syphilis 
 ignota or occulta is familiar to syphilologists and 
 dermatologists. 
 
 Some noteworthy investigations on this point 
 have been made by Hirschl, the results of which 
 are worth quoting. He found that, of 63 patients 
 sufferino- from o-ummatous affections of indubitable 
 syphilitic origin, in only 54 per cent, was there a 
 certain history of syphilis ; in 9*5 per cent, it was 
 probable ; while in 36"6 per cent, no evidence, other 
 than the patient's actual state, was forthcoming of 
 syphilis, hereditary or acquired. Practically the 
 same results were obtained by him in reference to 
 general paralysis. 
 
 Similar statistics to those obtained by him in 
 reference to syphilis and tabes have been quoted by 
 Gowers as having been arrived at by Dr. Pernet of 
 University College Hospital in reference to syphilitic 
 affections of the skin. Syphilis could only be 
 proved in 80 per cent, of the cases. 
 
 One may therefore say with justice that the 
 syphilitic origin of tabes and general paralysis is
 
 ON TABES DORSALIS 5 1 
 
 Statistically just as well proved as that of gummatous 
 affections of the skin. 
 
 Apart from juvenile tabes, which is always 
 syphilitic, hereditary or acquired, and which affects 
 both sexes in the same proportion, tabes occurs most 
 commonly among men of from 30 to 40 years of age, 
 and on the average from seven to ten years after the 
 primary syphilitic affection. There are, however, 
 exceptions to this rule, and I have seen it develop 
 in a man of yo, some forty to fifty years after 
 infection with syphilis ; and there is a case on record 
 (Berger) of a man of 74 becoming tabetic after 
 syphilitic infection at the age of 70. 
 
 Tabes at from 30 to 40 is a significant harvest of 
 the wild oats sown some ten years previously, and 
 the classes who sow wild oats most freely are those 
 who reap tabes most abundantly. Townspeople 
 furnish more cases of tabes than those who live in 
 the country ; and great centres of population with 
 their greater risks and temptations more than the 
 smaller. Tabes and general paralysis are rare 
 among priests, and of these only such as have had 
 syphilis, or exposed themselves to it (Bouchard), 
 
 Tabes is less common among women of the better 
 class than among those of the lower — doubtless, as 
 Moebius has su^o-ested, owing- to the fact that men 
 of the better class less frequently marry while still 
 suffering from infective syphilis. 
 
 But conjugal tabes is not uncommon, and most 
 neurologists have seen examples. Among others, 
 Mott has given several interesting illustrations.
 
 52 THE LUMLEIAN LECTURES 
 
 Almost without exception there has been syphiHtic 
 infection (90 to 96 per cent., Hudovernig) on one 
 side or the other ; usually the husband has infected 
 the wife. Nonne has related a very remarkable 
 family history. A child of 5 was infected with 
 syphilis through sleeping with a syphilitic bedfellow. 
 At 18 he became tabetic. The child infected the 
 mother, and the mother the father. Both these 
 became tabetic. In another family, described by 
 Souques (quoted by Erb), the father was syphilitic 
 and became a general paralytic ; the mother became 
 tabetic, and two young daughters also showed signs 
 of tabes — four in one family. Many similar cases 
 might be quoted in detail, but it is unnecessary to 
 multiply them. Syphilis, acquired or hereditary, 
 was the common factor in all. 
 
 And the question has been mooted (Lavalle. 
 Moebius, Erb) whether there may not be a form 
 of the syphilitic virus falling with special incidence 
 on the spinal and cerebral centres, and leading in 
 due course to tabes or general paralysis. Thus 
 Brosius relates the remarkable history of seven 
 glass-blowers who were infected in the mouth from 
 the same source in 1891. All were treated for 
 syphilis and dismissed as cured after a few weeks 
 of anti-specific treatment. In 1902, that is, after 
 eleven years, five of them came under his observa- 
 tion again. One was tabetic ; the second had 
 signs of general paralysis ; the third was already a 
 marked general paralytic ; the fourth had Argyll- 
 Robertson pupils, and indications of so-called
 
 ON TABES DORSALIS 55 
 
 tabo-paralysis ; while the fifth, at the time of 
 examination, was still in good health. The other 
 two had not been traced. The only common factor 
 in all these cases was the syphilitic infection. 
 
 It has been shown by Krafft-Ebing that general 
 paralytics are insusceptible to inoculation by the 
 syphilitic virus. This is a negative fact of great 
 significance, and of much more importance than if 
 the converse were true, for re-infection is always a 
 possibility. Whether the same holds in respect to 
 tabetics has not, so far as I know, been yet 
 demonstrated, but it is more than probable. Who 
 has ever seen a tabetic with a primary sore, or 
 even with secondary symptoms ? A considerable 
 number, both during life and after death, exhibit 
 tertiary manifestations, or traces of antecedent 
 syphilitic lesions. But these date from the infection 
 which preceded the tabes. Yet it cannot be said 
 that tabetics do not run the risk of re-infection, for 
 it is notorious that in the pre-ataxic stage tabetics 
 frequently exhibit excessive erotic propensities, and 
 expose themselves freely to the possibility of 
 acquiring syphilis. 
 
 Female children apart, tabetic virgines iiitactce are 
 rarcB aves. They are, however, not unknown. Of 
 two of whom I have read, the one (reported by 
 Nonne) was a tabetic girl of twenty, infected in 
 infancy by her mother suckling a syphilitic foster- 
 child at the same time ; the other (described by 
 Erb), a virgin of 2)^, who, some time previously, 
 had had a chancre on her lip.
 
 54 THE LUMLEIAN LECTURES 
 
 One might multiply arguments in favour ot the 
 causal relation between syphilis and tabes, but they 
 are unnecessary. For those above related, singly 
 and collectively, leave, in my opinion, little room 
 for doubt that tabes and general paralysis are in all 
 cases syphilitic, and that tshes pe?' se is as much a 
 proof of syphilis as a gumma of the skin. Those who 
 may be still unconvinced, and seek a more positive 
 sign, will probably have to wait till in sonie patho- 
 logical laboratory a monkey will be exhibited, 
 inoculated with a virus of syphilitic origin (a 
 product perhaps of the spirocluz'ta pallida), showing- 
 Argyll- Robertson pupils and loss of knee-jerks ; 
 and the question of the syphilitic origin of tabes 
 will be solved by the walk of a tabetic chimpanzee ! 
 
 It has been urged by v. Leyden and Goldscheider 
 as an argument against the syphilitic origin of tabes, 
 that the disease is rare among prostitutes, among 
 whom, nevertheless, syphilis is rife. But it must 
 be remembered that tabes is less common among 
 women than men ; and one must also take into 
 account the fact that as tabes is a late manifestation 
 of syphilis, it is not to be expected among those 
 who are still young enough to ply their wretched 
 calling. This accords with the statistics of Kron, 
 who, though he found no case of tabes among 148 
 young prostitutes, yet among 36 prostitutes over 
 25 years of age found 5 — a large percentage 
 enough. 
 
 Similarly it has been contended that in certain 
 countries and among certain races tabes is rare,
 
 ON TABES DORSALIS 
 
 55 
 
 though syphih's is common. Thus it has been 
 asserted that in Bosnia and Herzegovina, in Japan, 
 in Abyssinia, among the Arabs, and among negroes, 
 tabes does not occur. But more accurate investi- 
 gations have proved this assertion to be unfounded. 
 For tabes is common enough in Japan (Nose) ; it 
 is not unknown in Bosnia and Herzegovina ; it is 
 as common in Abyssinia as in Vienna (v. Halban) ; 
 and it certainly occurs among negroes (Collins, 
 Hecht). 
 
 It may not be so common among primitive races 
 as among those more advanced in civilisation and 
 all that this connotes, but this is not the point. 
 The fact that tabes only occurs where infection 
 with syphilis is possible, is the essential point in 
 this relation. 
 
 If tabes can be shown to exist, or to have existed, 
 in any land or people among whom syphilis has 
 never entered, and who have never had com- 
 munication with any outside syphilised world, the 
 exclusively syphilitic origin of tabes will have to be 
 abandoned, but not till then. 
 
 The relative infrequency of tabes among those 
 who have suffered from syphilis — it has been 
 variously calculated at from i to 5 per cent, of the 
 whole — is no argument against the syphilitic origin 
 of tabes. For there are many affections due to 
 syphilis, such as aortitis, hepatitis and orchitis, 
 which are relatively rare. And as Erb remarks, 
 the fact that perhaps only 1 in 10,000 of those 
 infected with tubercle suffers from Addison's disease
 
 56 THE LUMLEIAN LECTURES 
 
 does not disprove the tuberculous nature of the 
 affection of the suprarenal capsules. But the 
 smallness of the percentage, as well as other con- 
 siderations, suggest that, in addition to syphilis, 
 there may be some predisposing or other co- 
 operating causes determining the incidence of the 
 syphilitic virus on the nervous system, and the 
 development of tabes either of the brain or spinal 
 cord, or both. 
 
 Heredity, as such, plays no part in the etiology 
 of tabes, though, as we have seen, juvenile tabes 
 indicates a syphilitic inheritance. 
 
 Mention has already been made of sexual excess 
 and cold, to which great, but as we have seen undue, 
 prominence has been assigned by certain patho- 
 logists. The same may be said of neuropathic 
 inheritance, to which great importance was attached 
 by Charcot. It is not more frequent in tabes than 
 in other forms of nervous disease. Some facts, 
 however, suggest that certain families are specially 
 prone to develop tabes ; and several instances might 
 be quoted of more than one member of the same 
 family becoming the subject either of tabes or of 
 general paralysis, after acquiring syphilis. 
 
 Certain malformations, or developmental anoma- 
 lies, such as heteropia spinalis, have been recorded 
 by several observers (Mott, Orr and Rows, Rebizzi) 
 in the spinal cord of those affected by tabes or 
 general paralysis. All these have, no doubt, their 
 significance, but they are by no means constant nor 
 are they even common.
 
 ON TABES DORSALIS 57 
 
 The view that the substratum oi tabes Hes in 
 some defect of development, or vice of inheritance, 
 which predisposes to premature decay of the essen- 
 tial elements of a highly specialised tissue, and 
 reversion to those of a lower type, so that certain 
 individuals are, as it were, predestined to tabes, 
 reaches its culmination in Benedikt's burlesque, 
 Tabicus nascitui- : iion Jit ! No one who has had 
 experience of tabes in all ranks, and has seen 
 among its victims numbers of men of the most 
 healthy stock and vigorous physique, can regard it 
 otherwise than a gross libel to speak of tabetics in 
 general as being " connective-tissue individuals " 
 (Bindegewebs-individuen) or "degenerates" in any 
 true sense of the term. 
 
 Among other exciting causes of tabes, injuries 
 have been blamed, and many cases of so-called 
 traumatic tabes have been put on record. There 
 is no doubt that in certain instances — perhaps 5 
 per cent. — tabetic symptoms have first made them- 
 selves manifest after injuries of the spine or other 
 parts of the body, and the limb injured has not 
 infrequently been the primary, or principal, seat of 
 the characteristic pains : but a critical examina- 
 tion of such cases by Hitzig and others has failed 
 to substantiate the efficacy of trauma as a cause 
 of tabes independently of antecedent syphilitic 
 affection. 
 
 Of greater importance, perhaps, than any ot the 
 above-mentioned exciting or co-operating causes of 
 the tabetic deoeneration is the influence of over-
 
 58 THE LUMLEIAN LECTURES 
 
 Strain, or of over-exertion in debilitated states of 
 the system. The principle underlying these degen- 
 erations is that the normal equilibrium between 
 waste and repair which characterises health is dis- 
 turbed, and that in consequence of the process of 
 waste exceeding that of repair the neurons, especi- 
 ally those most engaged, tend to break down and 
 show dystrophic change. Edinger and Helbing 
 found that in rats artificially rendered anaemic, 
 which had been compelled to work in a treadmill, 
 or, by other device, to engage in long-con.tinued 
 muscular exertion, the posterior roots and posterior 
 columns of the spinal cord exhibited degenerations 
 similar to those of tabes. 
 
 These facts are suggestive, and appear to har- 
 monise well with the views of those who regard 
 fatigue and over- exertion as the most important 
 causes of tabes. As co-operative causes certainly, 
 but not apart from syphilis ; for the same degree of 
 fatigue and over-exertion which may excite tabes 
 in one in whom the soil is already prepared by 
 syphilis, may be borne by thousands of others 
 without material damage. 
 
 Edinger's "compensation theory" (Ersatz- 
 theorie) seems also to afford a plausible explanation 
 of the incidence of the affection in the parts most 
 subjected to continuous strain, the most common 
 being the sensory apparatus concerned in co-ordi- 
 nated locomotion and maintenance of equilibrium. 
 Occasionally, however, as in cervical tabes, the 
 arms, for similar reasons, are more especially
 
 ON TABES DORSALIS 59 
 
 affected ; and the pupils and optic nerves would 
 in like manner probably suffer more readily than 
 any of the other organs of special sense. 
 
 So also muscular atrophies would be more prone 
 to occur in the muscles which are most taxed, 
 similar to the occupation-muscular atrophies of 
 blacksmiths, railway porters, and the like. Pro- 
 bably the same principle may serve to explain the 
 incidence of the tabetic degeneration in one case 
 on the brain, and in another on the spinal cord. 
 Raymond has related an interesting narrative of 
 two brothers, similarly constituted by inheritance, 
 who both became infected by syphilis about the 
 same age. The one who lived a fast life and 
 indulged in sexual excess became a tabetic ; the 
 other who exhausted his brain in the struggle for 
 existence, became a general paralytic. Facts like 
 these are of eloquent significance. 
 
 The sensory neurones and posterior columns of 
 the spinal cord appear to be specially vulnerable to 
 certain toxic agents. Whether this can be explained 
 on evolutional or developmental principles ; or more 
 mechanically by peculiarities in the lymph circula- 
 tion in the posterior roots and posterior columns ; 
 or by an elective affinity of certain poisons for 
 special structures, such as nicotine for the cells of 
 the sympathetic ganglia, is a question for the future. 
 But degenerations similar to those of tabes have been 
 shown to occur from poisons introduced from with- 
 out or generated within, as well as in the course 
 of cerebral tumours, in some more obscure manner.
 
 6o THE LUMLEIAN LECTURES 
 
 In the great epidemic of ergotism which occurred 
 in Hesse in 1880, Tuczek observed many cases of 
 so-called ergotin tabes, in which, following the pre- 
 liminary acute gastric and other constitutional dis- 
 turbances, paraisthesise and ataxic symptoms set 
 in, sometimes several months after the ingestion of 
 the poison. Occasionally there was a tendency to 
 relapse, even when no further use was made of the 
 poisonous grain in the dietary. These ataxic and 
 sensory symptoms were found to coincide with 
 degenerative changes in the posterior roots and 
 posterior columns similar to those of tabes, 
 differing only in their acute course and in the 
 absence of sclerotic shrinkage. Lissauer's zones, 
 however, remained intact, and there were no pupil- 
 lary symptoms. 
 
 Somewhat similar degenerations, but less strictly 
 confined to the posterior columns, have been 
 observed in pellagra, endemic in certain districts, 
 especially in Northern Italy, and clue to the con- 
 sumption of poisonous maize. The degenerations 
 in this case affect not only the posterior columns, 
 particularly that of Goll, but also the antero-lateral 
 and pyramidal tracts. The exact nature and locali- 
 sation of these degenerations are, however, not as 
 yet precisely determined. 
 
 So also in lathyrism, caused by the seeds of the 
 lathy rus sativus, degenerations occur in the spinal 
 cord ; and a large number of inorganic poisons, 
 which specially cause peripheral neuritis, also occa- 
 sionally lead to degeneration of the posterior roots
 
 ON TABES DORSALIS 6 1 
 
 and posterior columns, such as alcohol, lead, arsenic, 
 and some others. 
 
 Many of the organisms which are the cause of 
 certain epidemic and endemic diseases, such as 
 those of diphtheria, beri-beri, and lepra, generate 
 toxins which not only induce multiple neuritis, but 
 also in some instances degeneration in the posterior 
 roots and in their intra-medullary processes and 
 collaterals. Whether these degenerations are 
 primary or secondary to the peripheral neuritic 
 processes, or whether the two are simultaneous, is 
 a question on which it is impossible at present to 
 pronounce definitely. Further — and perhaps of 
 greater significance — in reference to the pathogeny 
 of tabes, it has been shown that in certain cachectic 
 states of the system such as pernicious ancemia 
 (Lichtheim, Minnich, &c.), and diabetes (Fischer, 
 Williamson), probably by the autochthonous develop- 
 ment of toxins, degenerations occur in the spinal 
 cord, especially in the intramedullary processes of 
 the posterior roots, and tabetic-like symptoms ensue 
 in the form of parassthesiae, pains, and ataxic 
 disorders ; but however much these toxic affections 
 resemble tabes, they are in reality only pseudo- 
 tabetic, and none of them conforms in its clinical 
 history, course, and termination to tabes properly 
 so-called. Though weeks, or even months, may 
 elapse before the poisons introduced into the sys- 
 tem, or engendered in it by the specific organisms, 
 produce their deleterious effect on the nerves and 
 nerve centres, and though relapses may occur, there
 
 62 THE LUMLEIAN LECTURES 
 
 is a direct continuity between the introduction of 
 the poison and the development of the symptoms, 
 which run their course, and cease when the virus is 
 exhausted. In none of them do years elapse 
 between the original infection and the commence- 
 ment of the symptoms, and none has the progres- 
 sive character of true tabes. In none of them do 
 the children suffer like the juvenile tabetics, " who 
 have their teeth set on edge because their fathers 
 have eaten the sour grapes " of syphilis. 
 
 The similarity, however, between the degenera- 
 tions caused by these poisons and those of tabes, 
 strongly support the hypothesis, originally advanced 
 by Striimpell, that the tabetic degeneration is due 
 to the action of a toxin of syphilitic origin. The 
 tabetic degeneration is, as we have seen, a paren- 
 chymatous dystrophy, and has none of the features 
 usually considered characteristic of tertiary syphilitic 
 lesions. There are no cellular infiltrations ; no 
 specific vascular or gummatous changes ; and it does 
 not yield to the usual antispecific remedies. Hence 
 it has been customary to term the lesions of tabes 
 parasyphilitic (Fournier) or metasyphilitic (Moebius) 
 manifestations. It may be said, however, that this 
 is a mere petitio principii and that we have no 
 certain criterion of what constitutes a syphilitic 
 lesion. The spirochaeta pallida has not yet been 
 shown to be present in tertiary syphilis. But, in 
 any case, it would be incorrect to term a state of 
 abiotrophy (Gowers), or defective nutrition, a posi- 
 tive lesion. For this reason Lesser's term, quartar\-
 
 ON TABES DORSALIS 63 
 
 syphilis, however appropriate to certain late mani- 
 festations of syphilis in the liver and other organs, 
 is not strictly applicable to a condition in which 
 parenchymatous dystrophy precedes the develop- 
 ment of the sclerotic process. 
 
 The most probable pathogeny of the tabetic 
 degeneration, therefore, is that it is the result of a 
 toxin generated or conditioned by the syphilitic 
 virus. But this is at present a pure hypothesis, and 
 one which has to surmount great difficulties. Even 
 if, like the diphtherial neurotoxin, it may be the 
 product of the action of the specific organism on 
 the proteids of the body, or set free in the elabora- 
 tion of protective antitoxins, that it should take so 
 long to develop, or lie dormant for so many years 
 before exhibiting any signs of activity, is unlike 
 anything that we know of organ ismal toxins. There 
 is no parasyphilitic or metasyphilitic cerebro-spinal 
 lymphocytosis as such, apart from the commence- 
 ment of the tabetic process, and this is a fact of 
 great importance. Purves Stewart, who has made 
 investigations for me on this point in a series 
 of 12 cases, found that the average number of 
 lymphocytes in the centrifuge deposit of the cere- 
 brospinal fluid in secondary syphilis was 27, while 
 in tertiary syphilis it was o'9, or not more than the 
 normal. The tertiary cases with active cutaneous 
 lesions in progress had more lymphocytes than the 
 others, but even these were less than the average 
 of the secondary cases. The lowest count in tabes, 
 however, was 45. Most were between 80 and 100
 
 64 THE LUMLEIAN LECTURES 
 
 while the highest was 316 in the same relative 
 field. 
 
 Even if we were to consider it possible that there 
 might be long years of apparently normal health, in 
 spite of the presence of some malignant toxin in 
 the system, we have to reckon with the fact that 
 the tabetic process, once begun, is essentially of a 
 progressive nature, and postulates a more or less 
 continuous generation of the poison. 
 
 In this relation mention may be made of the 
 researches of Ford Robertson, McRae, and 
 Shennan. These observers found a bacillus, closely 
 resembling the Klebs-Loeffler bacillus of diph- 
 theria, in cultures from the walls of the respira- 
 tory, gastro-intestinal, and genito-urinary tracts of 
 patients dying from general paralysis, and from the 
 genito-urinary tracts of general paralytics and tabe- 
 tics durino- life. Rats fed on broth cultures of this 
 diphtheroid bacillus became ill after some weeks, 
 and when killed, showed in addition to signs of 
 Qfastro-enteritis, acute deg'enerative chanoes in the 
 cortex cerebri resembling to some extent those of 
 general paralysis, as well as degenerative changes 
 in the cells of the spinal cord. These observations 
 are of much interest, but a definite causal relation- 
 ship between this bacillus and tabes, or general 
 paralysis is far from being established. For the 
 same bacillus was found by them in the genito- 
 urinary tract in ten out of fourteen patients suffer- 
 ing from forms of insanity other than general 
 paralysis.
 
 ON TABES DORSALIS 65 
 
 I had written the above before Dr. Ford Robert- 
 son had delivered his Morisonian Lectures on The 
 Pathology of General Paralysis, now in course of 
 publication, and I am indebted to him for allowing 
 me to read the MS. of those as yet unpublished, 
 and to examine some of his microscopical prepara- 
 tions. He holds that general paralysis is the result 
 of a bacterial toxic infection from the respiratory 
 and alimentary tracts, owing to the impairment of 
 the resisting power of the organism by syphilis, 
 alcoholism, or the excessive use of nitrogenous 
 foods. The bacillus has the cultural and morpho- 
 loofical features, and reaction to Neisser's stain, of 
 the Klebs-Loeffler bacillus, of which it is an atten- 
 uated or more probably a special form, to which he 
 gives the nd,n\Q— bacillus paralyhcans . The bacil- 
 lus is polymorphic, and among other characters 
 which distinguish it from the xerosis bacillus and 
 Hoffmann's bacillus, is its power of forming acid 
 when cultivated in glucose broth. 
 
 The leucocytes of the blood of the general 
 paralytic, he maintains, have a greater dissolving 
 power over those bacilli than those of normal blood, 
 and the serum is also more lysogenic. 
 
 These diphtheroid bacilli, more or less altered 
 by lysogenic action, have been found by him and 
 his colleagues in the catarrhal pneumonic foci of 
 general paralytics, in the adventitial spaces of the 
 cerebral vessels, and in the meshes of the pia 
 arachnoid. They have also been found in the 
 blood, in the centrifuge of the cerebro-spinal fluid, 
 
 5
 
 66 THE LUMLEIAN LECTURES 
 
 and in the urine of the general paralytics, especially 
 during congestive attacks. From these fluids 
 cultures of the bacilli have been obtained, when 
 sufficient time has been allowed to destroy the 
 phagocytic or lysogenic action of the blood. Rats 
 infected or fed with cultures of these bacilli showed 
 changes in the cerebral cortex suggestive of those 
 of general paralysis, and similar appearances were 
 also found in a goat which had ingested them. 
 
 Robertson considers that in general paralysis 
 there is a life and death struofale between the leu- 
 cocytes and the diphtheroid bacilli, in which the 
 leucocytes succumb. There is a bacillary inva- 
 sion. Most of those which reach the blood are 
 digested, but many escape from the circulation, 
 either through the capillaries of the kidneys, or 
 through the walls of the cerebral blood-vessels 
 into the adventitial lymph channels. The further 
 disintegration of the bacilli in these lymph channels 
 gives rise to a local toxic action in addition to 
 the general toxaemia. The special cause of the 
 cord lesion in tabes is probably the absorption of 
 toxins from some peripheral septic focus such as 
 the bladder. 
 
 Interesting as these views are, there are many 
 difficulties in the way of their acceptance as a 
 satisfactory pathogeny of tabes, cerebral or spinal. 
 Eyre and Flashman, as the result of examination 
 durinor life of the throats of large numbers of 
 patients suffering from different forms of insanity, 
 as well as of the tissues of a considerable number
 
 ON TABES DORSALIS 67 
 
 of cases post mortem, could not determine any 
 greater incidence of diphtheroid organisms in the 
 throats of the insane than of the sane, and found 
 no evidence that b. diphtherice is more common in 
 the throats of general paralytics than of other insane 
 patients. Of ten cases of general paralysis ex 
 amined post mortem, diphtheroid organisms were 
 found in the respiratory tract in four, and the same 
 number in twenty-six cases of other forms of 
 insanity. They were unable to trace any causal 
 connection between b. diphtherice and general 
 paralysis of the insane. 
 
 I have the authority of Dr. Bulloch, one of 
 our eminent bacteriologrists, for the followinor 
 remarks : — Diphtheroid organisms are ubiquitous 
 both in health and disease. They may be found 
 in the intestines, in the genito-urinary passages, 
 in the throat ; in the nasal secretions, and else- 
 where in perfectly normal people, as well as in 
 catarrhal conditions of the respiratory tracts. The 
 characters given by Dr. Ford Robertson are not 
 sufficient to distinguish his bacillus paralyticans 
 from other forms of diphtheroid organisms. The 
 formation of acid in glucose broth is not special 
 to the bacillus paralyticans, for this is possessed in 
 different degrees by various other diphtheroids. 
 
 If diphtheroid bacilli can be found in the blood 
 films of living general paralytics, then general 
 paralysis must be regarded not as a toxaemia,, but 
 as a form of septicaemia, ranking with relapsing 
 fever as regards the dissemination of the organisms.
 
 68 THE LUMLEIAN LECTURES 
 
 If diphtheroid bacilH exist in the blood or cerebro- 
 spinal fiuid of general paralysis, they should be 
 capable of cultivation at once without waiting for 
 the cessation of any supposed bactericidal action, 
 for no lysogenic action of the serum for bacilli of 
 this group has been proved to exist. 
 
 Specimens of the cerebro-spinal fluid, blood, and 
 urine of several cases of tabes and general paralysis 
 have been cultivated for me by Dr. Emery of 
 King's College Hospital, and by Dr. Bulloch, and 
 have proved absolutely sterile. Dr. Ford Robert- 
 son emphasises the fact that he finds bacilli particu- 
 larly after so-called congestive attacks. It is not 
 improbable, therefore, that the bacilli which he has 
 found so largely diffused throughout the tissues in 
 general paralysis are only the result of a terminal 
 invasion. 
 
 In order to establish any causal relation between 
 the bacillus paralyticans and tabes, it would be 
 necessary to show that a toxin prepared from these 
 bacilli can produce the symptoms and post-mortem 
 appearances characteristic of tabes, cerebral or 
 spinal. Such evidence Ford Robertson has not 
 produced. The experiments on rats related by him 
 are not convincing, for it has been found by other 
 bacteriologists that rats are almost immune to the 
 most virulent form of the Klebs-Loeffler bacillus. 
 In presence of these and other considerations, we 
 may, therefore, say that the diphtheroid-bacillary 
 orioin of the tabetic toxin is at least not proven. 
 
 In the absence of any living organism capable
 
 ON TABES DORSALIS 69 
 
 of generating- the tabetic toxin, it may be that 
 the syphilitic virus under certain conditions so 
 affects some viscus or gland that in time it 
 develops, and continues to elaborate, some toxic 
 internal secretion which exerts its noxious influ- 
 ence on the nervous system. 
 
 In view of the brilliant therapeutic triumphs of 
 the present day, it is not, I think, too optimistic 
 to hope that, with the progress of bio- chemistry, 
 we may yet discover the source of the tabetic 
 toxin, its nature, and the means of neutralising it 
 and staying its ravages. We cannot expect to 
 repair damage already done, for intramedullary 
 degeneration is irreparable, but we may at least 
 succeed in arresting the disease. This not infre- 
 quently happens spontaneously ; and the thera- 
 peutic measures last employed are apt to get the 
 credit of the happy result. But beyond palliation, 
 or the temporary relief of urgent symptoms, no 
 remedy, chemical or physical, has as yet been 
 discovered which materially influences the morbid 
 process. Antispecific treatment may be of service 
 against possible coincident tertiary lesions, or may 
 even cure pseudo-tabetic syphilitic symptoms ; but 
 certainly ordinary antispecific treatment exerts no 
 beneficial influence on true tabetic degeneration, 
 and may, if carried to extremes, prove positively 
 harmful. If, however, we believe, as I think we 
 must, that the syphilitic virus is the sine qua non 
 in the etiology of tabes, it ought to be our en- 
 deavour to combat this with all the resources at 
 our command.
 
 /O THE LUMLEIAN LECTURES 
 
 The comparative mildness of the primary con- 
 stitutional symptoms in those who ultimately be- 
 come tabetic, would almost seem to indicate that 
 when the syphilitic virus expends itself in severe 
 primary and secondary manifestations there is less 
 tendency to the subsequent development of the 
 subtle poison which proves so disastrous to the 
 nervous system. 
 
 This mildness of the primary symptoms with- 
 out doubt leads, in many cases, to insufficient 
 treatment. Even the most efficient antispecific 
 treatment, however, may fail to prevent the sub- 
 sequent development of tabes in certain cases. 
 But the facts recently brought by Fournier before 
 the Academie de Medecine in reference to the 
 relation between general paralysis and syphilis 
 and its treatment are worthy of careful consider- 
 ation and probation. 
 
 Fournier states that of seventy-nine cases of 
 general paralysis coming under his own observa- 
 tion, in which he was able to obtain precise data 
 as to the character and duration of the treatment 
 to which they had been subjected, only 5 per 
 cent, had been treated with thoroughness ; while 
 in at least 80 per cent, the antispecific treatment 
 had been carried out in a more or less perfunctory 
 manner. It is not enough, he says, to treat 
 constitutional syphilis for weeks or months only. 
 Fournier contends that mercurial treatment should 
 be administered for at least two years continuously, 
 and intermittently for several years longer. " Treat
 
 ON TABES DORSALIS 7 1 
 
 syphilis well, and you will have few cases of 
 general paralysis," he says. The same would 
 hold true of spinal tabes. These views deserve 
 the special consideration of those who have most 
 opportunities of treating syphilis in its earlier 
 stages. If they are substantiated tabes, both 
 cerebral and spinal, will become one of the rarer, 
 instead of one of the most common, and, it is 
 to be feared, increasing forms of organic nervous 
 disease.
 
 72 THE LUMLEIAN LECTURES 
 
 Lecture III. 
 
 Physiological Pathology of Tabes. 
 
 I must now pass from the realms of speculation 
 as to the nature of the tabetic toxin and the thera- 
 peutic possibilities of the future to a more abstruse 
 part of my subject — namely, to attempt to corre- 
 late the essential symptoms of tabes with their 
 anatomical and physiological substrata. For then 
 only can we be said to have constructed a complete 
 pathology of a disease when we are able to refer 
 each symptom to its organic basis. To do this, 
 however, in the case of such a polymorphic disease 
 as tabes is a vast undertaking, and involves a wide 
 and deep survey of the relations of the nervous 
 system to almost every function of the organism — 
 somatic, visceral, and trophic. For no disease ex- 
 hibits such a complex sym.ptomatology or puts to 
 us such a host of neuropathological problems. On 
 many of these problems I should despair of being 
 able to say anything worthy of your attention which 
 has not been said, and perhaps better said, by 
 others before me. I purpose, therefore, to limit 
 my survey of the physiological pathology of tabes 
 to a few only of its more prominent symptoms, in 
 respect to which clinical and experimental research 
 have lately been busy and more or less suggestive 
 or fruitful in results.
 
 ON TABES DORSALIS J T) 
 
 Of all the symptoms of tabes, ataxy is the most 
 characteristic. All tabetics, however, are not 
 ataxic, for some become blind from optic atrophy, 
 or suffer for years with excruciating lightning pains, 
 bladder and other visceral troubles, without becom- 
 ing ataxic. Some neuro-pathologists, with strange 
 logical incoherence, regard the occurrence of optic 
 atrophy as exerting a kind of inhibitory influence 
 on the development of ataxic symptoms. All that 
 is justified is that the ataxic and amaurotic forms 
 of tabes are often more or less distinct, though not 
 exclusive of each other. Tabes is, as we have seen, 
 essentially an affection of the sensory spinal pro- 
 toneurone, whether in its ganglionic cell, peripheral, 
 or intramedullary process, and in this we have to 
 seek for an explanation of its most important 
 symptom. For all other changes, whether in 
 the peripheral motor nerves, anterior cornua, or 
 bulbar nuclei, molecular layer of the cerebellum 
 (Weigert), tangential fibres of the cortex cerebri 
 (Jendrassik), or large cells of the post-central gyrus 
 (Campbell) and their processes, are either incon- 
 stant, or mere concomitants, if not the results of 
 the primary sensory degeneration. The motor 
 neurone is intact ; the electrical reactions of the 
 muscles are normal, and their dynamic strength 
 not necessarily impaired. The paths of impulse 
 from the cortical centres are free from degenera- 
 tion ; and so also are the efferent tracts from the 
 cerebellum by way of Deiters' nucleus, as well as 
 the quadrigemino-spinal and rubro-spinal antero-
 
 74 THE LUMLEIAN LECTURES 
 
 lateral tracts from the mesencephalon. The lesion 
 is therefore on the afferent or sensory side of the 
 apparatus of movement. If it is possible to give 
 a satisfactory explanation of tabetic ataxy by the 
 demonstrable lesion of the centripetal paths to the 
 spinal, subcortical (mesencephalic and cerebellar), 
 and cortical centres, it would seem unnecessary, 
 with Erb and others, to ascribe the ataxy to lesion 
 of some hypothetical centrifugal tracts of motor 
 co-ordination. 
 
 It is more than probable that certain ataxic, or 
 better, astatic muscular disorders may arise from 
 anomalies of excitation, or irregularities in the 
 transmission, of impulses from the motor centres, 
 such as are exemplified in multiple sclerosis, lesions 
 of the tegmentum, and some other intracranial 
 lesions. But these differ in all their essential 
 features from tabetic ataxy, and therefore need not 
 occupy our attention further in reference to the 
 question before us. 
 
 The degeneration of the centripetal tracts in tabes 
 manifests itself in manifold abnormalities : — per- 
 version, retardation, diminution, or total loss of 
 sensory conduction in the domain of epicritic (tactile 
 proper), protopathic (algesic, thermal), and deep 
 sensibility, including all those factors — muscular, 
 tendinous, articular — which form the basis of the 
 sense of pressure and the complex muscular sense 
 (Head). 
 
 Associated with this from the commencement, 
 and it may be throughout the whole course of the
 
 ON TABES DORSALIS 75 
 
 malady, and dependent on some subtle bio-chemical 
 or bio-physical processes into the arcana of which 
 we have not yet been able to penetrate, are those 
 repeated and violent storms, somatic and visceral, 
 which torture and rack the unhappy patient. 
 Whether these lightning pains and crises depend 
 on changes in the peripheral nerves, as some ob- 
 servers {Gowers, &c.), are inclined to believe, or 
 have their basis in the central processes of the 
 spinal ganglion, is uncertain and difficult to deter- 
 mine ; but the fact that many of the subjective 
 paraesthesiae (such as the girdle sensation) and the 
 objective signs of sensory impairment are to a large 
 extent of a segmental or root distribution is, inter 
 alia, in favour rather of the root origin of the tabetic 
 neuralgia. And the fact that peripheral neuritis is 
 relatively much less common in tabes than the 
 lightning pains points in the same direction. 
 
 The impairment of cutaneous sensibility— epicritic 
 or protopathic — varies enormously in frequency, 
 degree, and range, and there is no constant rela- 
 tion between the ataxic symptoms and the extent 
 to which cutaneous sensibility is affected. It may 
 even happen that there is no obvious affection of 
 cutaneous sensibility to the most delicate tests. 
 This is a valid argument against the sensory theory 
 of ataxy as first propounded by v. Leyden, who at 
 one time put too much stress on the impairment 
 of cutaneous sensation. The impairment of cuta- 
 neous sensibility is more marked in its protopathic 
 (algesic) than in its epicritic (tactile) form, and is
 
 76 THE LUMLEIAN LECTURES 
 
 most common below the knees. Roughly speaking, 
 the incidence of tactile (hyp-) anaesthesia corre- 
 sponds with that of (hyp-) analgesia, though the 
 percentage of the latter is much greater than that 
 of the former.* 
 
 It is quite otherwise, however, in respect to the 
 relations between ataxy and impairment of the deep 
 sensibility, which, as has been shown by Sherrington 
 and Head, is dependent on the nerves which run 
 with the motor nerves and originate in Pacini- 
 like organs (muscle-spindles, Golgi organs, &c.), 
 and are specially adapted to be influenced by pres- 
 sure, muscular contraction, and change of position. 
 A third to a half of the nerves which are distributed 
 to muscles are of this afferent type (Sherrington). 
 
 Affections of this system of nerves reveal them- 
 selves subjectively in various ways, such as by 
 insensibility to passive movements of the joints, 
 inability to determine the exact position of the 
 limbs or their segments, and defective sense of 
 muscular contraction; these are individually affected 
 with great diversity in different cases. 
 
 Frenkel holds that in every case of ataxy careful 
 investigation, by precise methods calculated to 
 eliminate all active muscular movements on the 
 part of the patient, never fails to reveal a more or 
 less extensive impairment of the sense of passive 
 
 * I am making these general statements on the basis of a 
 careful investigation of the affections of sensibility in fifty 
 cases of tabes by Dr. Golla at the National Hospital for 
 Paralysed and Epileptic, who has kindly placed his results 
 at my disposal.
 
 ON TABES DORSALIS "] '] 
 
 movements of the articulations ; and that this is 
 more pronounced and more widespread in the more 
 ataxic Hmb, when, as is not infrequently the case, 
 they are unequally affected. The loss of the sense 
 of passive movements is, according to Friedlander, 
 always manifested first in the toe-joints, and tends 
 to spread upwards to the higher joints with the 
 increase of the ataxy. But the results of Golla's 
 investigations at the National Hospital would lead 
 me to question the accuracy of Frenkel's statements; 
 for in a considerable percentage of cases of obvious 
 ataxy he was unable to obtain any evidence of 
 impairment of the sense of passive movements. 
 And though it is the rule that this sense is more 
 impaired in the more ataxic limb in the same 
 individual, the rule is by no means absolute (Fried- 
 lander). And it is a remarkable fact that the ataxic 
 disorders vary enormously in different patients in 
 whom, to all appearances, the affections of super- 
 ficial and deep sensibility are alike. 
 
 The muscles of the ataxic limbs are often tolerant 
 of the severest pressure without discomfort. Closely 
 related to this is the fact to which my attention 
 has been called by Golla, and which I have fre- 
 quently verified — namely, that the sense of tonic 
 or tetanic muscular contraction induced by the 
 faradic current is greatly diminished or totally lost, 
 while the sense of minimal or rapid contraction 
 is well preserved. This is at once capable of 
 demonstration in cases in which there is already 
 complete cutaneous anaesthesia, and can in others
 
 78 THE LUMLEIAN LECTURES 
 
 be shown after preliminary kataphoresis of cocaine 
 into the skin. 
 
 Not infrequently both forms of musculo-con- 
 tractile sensibility are diminished or lost, but the 
 two stand in no constant relation to each other ; 
 the sense of tonic contraction being by far the 
 most frequently affected.* 
 
 Golla has luminously suggested that the loss of 
 the sense of tonic contraction is the sensory side 
 
 * Dr. Golla has furnished me with the following statement : 
 In two cases there was no obvious ataxy of the arms or hypo- 
 tonia, while the legs were markedly hypotonic, and completely 
 anaesthetic and algesic, with nearly complete loss of the sense 
 of position. In both these cases there was only very slight 
 difference in the acuity of the sense of rapid muscular contrac- 
 tion ; while both cases had lost the sense of tonic contraction of 
 the legs, and preserved it little, if at all, affected in the arms. 
 I regard these two cases as of cardinal importance, as we 
 were able to compare limbs deeply affected with limbs hardly 
 at all affected, either as regards ataxy or any other form of 
 sensation or hypotonia. In seven other cases there was in all 
 more or less marked depression of the sense of tonic contraction 
 in the less affected limbs, and complete absence of that sense 
 in the limbs in which hypotonia and ataxy were well marked. 
 In none of these cases was there any affection of the sense 
 of rapid muscular contraction with the exception of two, 
 where all forms of muscular sensation were completely lost. 
 That the sense of rapid contraction of the muscle does not 
 depend on the cutaneous sensibility is shown in three cases, 
 where it was preserved with complete cutaneous anaesthesia. 
 That the absence of the sense of tonic contraction is not a 
 cause of loss of sense of position, but rather a specific pheno- 
 menon of tabes associated with the loss of the reflex tonus 
 and depression of the spinal reflexes, is suggested by two 
 cases of ataxy following neuritis, where it was preserved, 
 though there was complete absence of the sense of position 
 and of passive movement. In these cases tonicity had 
 returned to the affected limbs.
 
 ON TABES UORSALIS 79 
 
 of the loss of reflex tonus which seems the lead- 
 ing feature of tabes, while the sense of quick con- 
 traction may be the sensory analogue of the quickly 
 contractinor elements of the muscle. 
 
 This latter suggestion would be in harmony with 
 the contractile nature of the muscle spindles, which, 
 as Sherrington has suggested, may function in 
 signalling active rather than passive movements. 
 
 Another form of sensibility which is frequently 
 impaired or lost in tabes is the sense of vibration 
 (pallsesthesia) of a tuning-fork, first described by 
 Egger (1899), s^rid attributed by him to the bones 
 or periosteum. But it can be perceived in parts 
 where there are no bones (Rydel and Seiffer), and 
 the continuity of bone is not necessary to its 
 propagation (Minor). It is probably dependent 
 principally on the deep sensibility which forms the 
 basis of the perception of passive movements and 
 the position of the limbs, but it does not vary 
 exactly with this (Dejerine, Marinesco, self) and 
 its influence in respect to motor co-ordination is 
 still sub jiLciice.'^ 
 
 Of greater constancy, and therefore of greater 
 fundamental importance in the pathology of ataxy, 
 is the impairment or total loss of those centripetal 
 impressions which, arising chiefly in the muscles 
 themselves and acting on the spinal and subcortical 
 
 * Dr. Head informs me that, in his opinion, it may be 
 partly due to the rapidly-repeated single shocks acting on 
 the epicritic nervous system, and therefore akin to the appre- 
 ciation of simultaneous compass points.
 
 So THE LUMLEIAN LECTURES 
 
 (cerebellar) centres, are the basis of reflex muscular 
 tone. This is shown in the loss of the so-called, 
 but falsely so-called, tendon reflexes, of which the 
 best known (since Westphal first called attention 
 to it in 1875) and most commonly tested is the 
 knee-jerk ; though the Achilles-jerk is of equal 
 or even greater importance. Others also, usually 
 obtainable in normal conditions, disappear as ataxy 
 becomes more general. Associated with this is 
 that remarkable condition of the muscles to which 
 Frenkel has given the appropriate name hypotonia. 
 This condition, so general in ataxy, had not 
 altogether escaped the notice of v. Leyden, but 
 it is to Frenkel that we owe a more precise know- 
 ledge of its characters and significance. The 
 muscles are not necessarily flaccid to palpation, 
 but they permit an undue mobility of the limbs 
 and trunk, and do not develop resistance as an- 
 tagonists when put on the strain like normal 
 muscles. To this condition many of the charac- 
 teristic attitudes of the ataxic, recumbent or erect, 
 and the brusque overaction of the prime movers 
 of the limbs in walking, are largely due. Better 
 than verbal description of this condition are the 
 accompanying illustrations (figs. 29 to 36). 
 
 This condition is due, as we have said, mainly 
 to the loss of the tonic action of the spinal centres, 
 but it may in part also be dependent on failure 
 of the similar tonic influence which Luciani 
 in animals, and Holmes and Grainger Stewart 
 in man, have shown to be exercised by the
 
 ON TABES DORSALIS 
 
 81 
 
 cerebellum. The loss of cerebellar tone, how- 
 ever, is not associated with loss of the tendon- 
 jerks, and may exist when these are normal, or 
 even increased. The non-correspondence between 
 
 Fig. 29. — Maximal passive extension of knee in a normal individual 
 (Frenkel). 
 
 Fig. 30. — Hypotonia of popliteal muscles in tabes (Frenkel). 
 
 the state of the knee-jerks after lesions of the 
 cerebellum and spinal lesions with loss of tone, 
 made me at one time question the existence of 
 Luciani's cerebellar atonia. I do not, however, 
 6
 
 82 
 
 THE LUMLEIAN LECTURES 
 
 doubt the reality of a condition to which perhaps 
 the term is not altogether inappropriate. 
 
 The basis of this loss of tone, and of the normal 
 antagonistic counter-tension to active and passive 
 
 Fig. 31. — Maximal flexion of hip with extended knee, in a normal 
 individual (Frenkel). 
 
 Fig. 32. — Moderate hypotonia of hamstring muscles in tabes 
 (Frenkel). 
 
 movements, is the degeneration of the reflex col- 
 laterals of the posterior roots to the anterior 
 cornua, and those to Clarke's columns, and thence 
 indirectlv to the cerebellum bv vvav of the direct
 
 ON TABES DORSALIS 
 
 83 
 
 cerebellar tract. By the long- fibres of the posterior 
 columns in some measure, and by other paths, the 
 position of which in the spinal cord is still the 
 subject of considerable diversity of opinion, and on 
 which the last word has not yet been said, all the 
 
 Fig. ^2- — Maximal hypotonia of hamstring muscles in tabes 
 (Frenkel). 
 
 Fig. 34. — Hypotonia of adductors in tabes (Purves Stewan). 
 
 conscious impressions, or sensations proper, derived 
 from the skin, muscles, and deep structures are con- 
 veyed to the cerebral centres or centres of per- 
 ception. To discuss these, however, at greater 
 length, would be beyond the scope of the subject
 
 84 THE LUMLEIAN LECTURES 
 
 in hand. And it is not as yet possible to say 
 for certain how much of the sensory loss in tabes 
 is due to the affection of the posterior roots them- 
 selves, and how much to that of the intraspinal 
 paths, or how much to both combined. 
 
 Tabetic ataxy is probably the complex result 
 of blocking of all the forms of sensory impression, 
 conscious as well as unconscious, more or less, and 
 not of one kind exclusively. 
 
 Fig. 35. — Hypotonia of gluteal and hamstring- muscles (Purves 
 Stewart). 
 
 Ataxy may result from lesion of the sensory 
 nerves alone, apart from any implication of the 
 spinal tracts. This is well exemplified in the 
 pseudo-tabes or iieiiro- tabes peripherica (Dejerine), 
 which is not an uncommon result of the toxic 
 action of alcohol, diphtheria and other peripheral 
 nerve poisons. 
 
 So far as the ataxic symptoms are concerned, 
 these cases are not disting-uishable from those of
 
 ON TABES DORSALIS 
 
 85 
 
 true tabes. The abnormalities of superficial and 
 deep sensibility are similar ; Romberg's symptom 
 and inco-ordination of the limbs, intensified by 
 shutting the eyes, are alike ; the knee-jerks are 
 absent, and though there are symptoms and signs 
 which enable a differential diagnosis to be made, 
 yet the two affections resemble each other so 
 closely that they can be readily mistaken tor each 
 other. 
 
 Fig. 36. — General hypotonia (Purves Stewart). 
 
 As to the effects on movement of experimental 
 section of the posterior roots, which means cutting 
 off centripetal impressions of all kinds, conscious 
 and unconscious, there has been some controversy 
 of late. In the first instance, at least, as has 
 been known since Bell's classical experiment of 
 section of the fifth nerve, and illustrated by Mott 
 and Sherrington in their experiments on the
 
 86 THE LUiMLEIAN LECTURES 
 
 posterior roots of the brachial plexus in monkeys, 
 there is an indirect paralysis — termed centripetal 
 paralysis by Hering — in the related parts. This, 
 however, as one might a priori expect from 
 clinical observation in man, should only continue 
 so long as the sensations from the anaesthetic 
 limb had not been replaced or compensated by 
 the eye. And in reality, as Munk, who has 
 repeated Mott and Sherrington's experiments, has 
 shown, even in monkeys the paralysis is only tem- 
 porary, and in course of time the animals learn 
 to use the anaesthetic limb again. But, and this 
 is the cardinal point, the movements of the limb 
 are never normal, but remain "brusque, excessive, 
 and awkward " {oj'iisk, iibermdssig, inigesckicki). 
 This, though somewhat w^anting in precision, is 
 a fair description of an ataxic limb. 
 
 And Bickel and Ewald found, after similar experi- 
 ments on doCTs, that the ataxic movements of the 
 limbs, which tended in course of time to diminish, 
 were reinduced when the animals were blindfolded 
 or placed in a dark room : — their limbs being ren- 
 dered visible by luminous paint. The resemblance 
 of these symptoms to those of ataxy in man is 
 remarkable. 
 
 The spinal paths of afferent impressions include 
 not only the reflex collaterals to the anterior cornua 
 and Clarke's columns, but also those of sensation 
 proper to the cortex cerebri. It would seem almost 
 impossible to dissociate these from each other, but 
 yet such a thing apparently occurred in one case
 
 ON TABES DORSALIS Sj 
 
 which has been put on record (Schilppel) so lono; 
 aoro as 1874. This was a case probably of syringo- 
 myelia, and though the symptoms and post-mortem 
 appearances have not been recorded with such ful- 
 ness as we should consider necessary at the present 
 day, yet it is one of great importance as showing 
 that total interruption of the spinal paths of con- 
 scious sensation does not of itself cause ataxy 
 proper. The case was that of a man who became 
 completely anctsthetic to tactile, thermal, and pain- 
 ful stimuli ; lost all sense of pressure, and had no 
 sense of active or passive movements of his limbs 
 or joints, and yet, under the guidance of vision, 
 could stand, walk, and execute all voluntary move- 
 ments in a fairly normal manner. The cord lesion 
 was a cavity occupying chiefly the grey matter, 
 especially the posterior cornua and grey commis- 
 sure. The anterior cornua were intact. Whether 
 in this case the reflex collaterals were spared we 
 cannot froni the description say positively, but it 
 is not improbable. 
 
 If we ascend higher towards the cerebrum we 
 have no evidence of organic lesion strictly limited 
 to the paths of sensation apart from those of 
 motion, either in the mesencephalon, internal cap- 
 sule and optic thalamus, or higher. And though 
 in such cases certain so-called hemiataxic disorders 
 of movement may be observed, probably, in part 
 at least, dependent on impaired transmission of 
 sensory stimuli, they are in many respects quite 
 unlike tabetic ataxy. The only conditions which
 
 88 THE LUMLEIAN LECTURES 
 
 enable us to eliminate the influence of true percep- 
 tion on the co-ordination of movements are, inter 
 alia, cases of functional anaesthesia. These are 
 usually hysterical and mostly unilateral, and the 
 symptoms are so equivocal and contradictory of 
 each other that they do not afford a safe basis on 
 which to build. It is sufficient to say that in 
 functional cerebral anaesthesia, notwithstanding" the 
 apparent total loss of all kinds of sensation, cuta- 
 neous and deep, co-ordinated movements of great 
 complexity may still be effected under the guidance 
 of vision, actual or ideal. That clear consciousness 
 of kinaesthetic impressions is unessential is proved 
 by the phenomena of sleep-walking and other states 
 in which the functions of the hiohest centres are 
 for the time in abeyance. 
 
 The general result of this analysis is, that tabetic 
 ataxy depends on impairment or loss of centripetal 
 impressions of all kinds, conscious as well as un- 
 conscious, and, of these, those from the deeper 
 structures (muscles, tendons, joints) are more im- 
 portant than the superficial ; and if one may single 
 out one class as more essential than the others of 
 the complex, I would specify those to the spinal, 
 subcortical and cerebellar centres. 
 
 Vision can more or less completely compensate 
 for loss of kinaesthetic sensation ; and largely also, 
 but to a much less extent, for the loss of the purely 
 afferent impressions to the spinal and cerebellar 
 centres. In the absence of vision, all the disorders 
 of co-ordination, static and dynamic, are greatly
 
 ON TABES DORSALIS 89 
 
 intensified, and this is the pathognomonic feature 
 of tabetic ataxy. 
 
 The constitution and connections of the spinal 
 and subcortical centres provide for the combined 
 action of the various muscles (prime - movers, 
 synergies, antagonists, and collaterals) concerned 
 in simple and co-ordinated acts — for example, the 
 synergic extension of the wrist in closure of the fist; 
 and the combined flexion of the thigh, leg, and 
 dorsiflexion of the foot of the swinging leg in 
 walking. 
 
 This is proved by the perfectly co-ordinated 
 movements which occur in decapitated animals in 
 response to various external and internal stimuli, 
 and by the similar phenomena that can be elicited 
 by stimulation of the posterior roots after transection 
 of the cord in monkeys (Page May). 
 
 "The afferent path evokes discharge only of 
 harmoniously related muscular groups. If it be 
 stimulated strongly the number of motor groups 
 excited may be extensive, and their situation in- 
 volve the whole length of the spinal axis ; but the 
 groups excited contemporaneously are always har- 
 moniously acting groups— for example, flexors of 
 hip with flexors of knee, and not groups mutually 
 subversive of each other's action. Each individual 
 muscle of each group, so far as analysed, seems 
 in its action balanced and subordinated as in a 
 " natural movement " (Sherrington). 
 
 The conscious accompaniments of muscular 
 action, both in the process of learning and through-
 
 90 TH1-: LUMLEIAN LECTURES 
 
 out life, play a more important I'tye in human 
 actions in all their manifold diversity than in the 
 more simple activnties of the lower animals, and 
 their removal — potential or real — exercises a greater 
 disturbance of the whole mechanism ; but the 
 mechanism is not formed by conscious activity- 
 it already exists. 
 
 It has been said by Forster that in the newborn 
 infant the synergic extension of the wrist on reflex 
 closure of the fist does not occur, and that it only 
 becomes developed as the result of cerebral selec- 
 tion. I should seriously question the accuracy of 
 this statement, for though it is difficult, as I have 
 found, to be certain of the simultaneous action of 
 the extensors of the wrist in the refiexly-excited 
 closure of the fist in newborn intants, I have occa- 
 sionally observed it, and Dr. Beevor informs me 
 he has done the same ; and it would be difficult 
 to explain the extraordinary power of grip pos- 
 sessed by newborn infants without assuming that 
 the extensors of the wrist were acting synergically 
 at the same time ; for infants only a few hours 
 old, as Louis Robinson has shown (" Darwinism 
 in the Nursery," Nineteenth Century, November, 
 1 891), can hang" on to the finger or a small stick, 
 and sustain their whole weight for several seconds 
 — a power which increases for several weeks after 
 birth. Certainly the synergic extension of the 
 wrist becomes manifest long before the child can 
 have developed it by conscious selection. 
 
 In volition, or action conditioned by conscious
 
 ON TABES DORSALIS 9I 
 
 discrimination, the cortical centres act throuo-h the 
 spinal, but in the cortex there is oreater differ- 
 entiation and power of dissociation than in the 
 spinal cord, so that individual movements, if not 
 individual muscles, can be isolated from their usual 
 synergists. Thus we can voluntarily flex the 
 fingers without extending the wrist. But our 
 powers in this respect are limited, and we can only 
 eliminate such svnero-ists and collaterals as we are 
 clearly conscious of; while others, to which our 
 attention has not been specially called, co-operate 
 as usual, and their action comes to us almost as 
 a revelation when it is pointed out. 
 
 How the cortical motor centres act un those of 
 the spinal cord, whether directly through the 
 pyramidal tracts, or indirectly through association 
 cells (Gad, v. Monakow, Beevor), or through the 
 same channels as the reflex collaterals (Schafer), 
 is at present a matter only of speculation. But if, 
 as Schafer contends, the fibres of the pyramidal 
 tracts arboresce not around the cells of the anterior 
 cornua but around those of Clarke's columns, and 
 at the base of the posterior cornua, then the hypo- 
 thesis that the cortical centres act through the same 
 channels as the reflex collaterals would be the 
 most probable. For the phenomena of ataxy and 
 pseudo- ataxy clearly demonstrate the necessity of 
 kinaesthetic impressions to secure the due co-opera- 
 tion of the synergic, antagonistic, and collateral 
 muscles with the prime movers in any given 
 movement.
 
 92 THE LUMLEIAX LECTURES 
 
 It would appear from the researches of Sher- 
 rington and Hering- that the innervation of a given 
 muscle ipso facto inhibits its true antagonist — 
 so-called I'-eciprocal innervation. This is well ex- 
 emplified in the case of paralysis of the internal 
 rectus of the one eye — say the left. On innerva- 
 tion of the lateral conjugate movement to the right 
 the divergent strabismus of the left eye diminishes, 
 and the eyeball swings up to the middle line, owing 
 to inhibition of the tone of the unantagonized 
 external rectus. 
 
 The simultaneous removal of all obstruction to 
 the unfettered action of a muscle, implied in the 
 law of reciprocal innervation, may be regarded as 
 a useful provision ; but it is a law subject to being 
 overruled by the varying necessities of the case 
 and the effects desired. Thus in sudden extension 
 of the forearm the flexors are inhibited ; but in 
 extension of the forearm against gravity the flexors 
 are obviously in action as well as the extensors. 
 And as Duchenne, Richer, Demeny, and others 
 have demonstrated, the action of the true antago- 
 nists is in manifold ways brought into play to 
 moderate and steady that of the prime movers. 
 
 The essence of ataxy, in all its manifestations, 
 is the loss of, or impaired transmission of the 
 kinsesthetic impressions to the spinal, subcortical, 
 and cerebral centres, on which the graduated and 
 harmonious co-operation of all the muscles con- 
 cerned in our various actions depends. 
 
 To analyse these in detail, however, would take
 
 ON TABES DORSALIS 93 
 
 a volume. Nothing could be more admirable in 
 this respect than the masterly analysis of tabetic 
 disorders given by Forster in his work on the 
 physiology and pathology of co-ordination. 
 
 One of the most characteristic features of tabetic 
 ataxy is the overaction of the prime movers. This 
 is manifested in almost every form of voluntary 
 movement. It is most marked when there is also 
 impairment of conscious sensation in the limbs. It 
 is, no doubt, largely due, and has been attributed 
 almost exclusively, to the defective sense of move- 
 ment achieved, so that the patient has no guide to 
 the inhibition of further motor innervation. Hence, 
 especially when his eyes are shut, he continues the 
 action beyond what is required, and is himself 
 astonished when he sees how far he has overshot 
 the mark. But this is not all, nor is it the most 
 important factor in the sudden and immoderate 
 character of the prime movements of the ataxic. 
 This consists in the hypotonia and default of 
 the afferent impressions which normally secure 
 the development of the reflex resistance of the 
 antagonists. The recumbent ataxic, therefore, 
 when told to lift his foot from the bed does so with 
 abnormal energy and brusqueness, because the 
 extensors of the thigh fail to develop the resistance 
 necessary to moderate the action of the flexors. 
 So also the collateral muscles fail to steady the 
 limb, so that it swerves from the direct line, or 
 oscillates in various directions about the mark aimed 
 at. The default of the centripetal spinal and sub-
 
 94 THE LUMLEIAN LECTURES 
 
 cortical impressions is exemplified in many other 
 ways. Even in the recumbent attitude the limbs 
 of the ataxic exhibit abnormalities of position. The 
 le^s are rotated outwards or inwards, from defective 
 reflex tone of the inward or outward rotators of the 
 thigh ; and in particular the foot assumes the 
 equinus or equino- varus position from defective 
 tone of the dorsiflexors. 
 
 When the leg is used as an instrument of pro- 
 gression, the " impulsive association," as Duchenne 
 expresses it, or the synergetic combination of the 
 flexors of the thigh, flexors of the leg and dorsi- 
 flexors of the foot, is dissociated, so that the 
 dorsiflexion of the foot fails to occur, and the toe 
 tends to drao- on the floor. This defect is corrected 
 under the guidance of the eye. but it is generally 
 overdone. So also when the foot is again planted 
 on the floor it not uncommonly becomes inverted 
 or everted, from failure, as Forster points out, of 
 the reflex tone either of the peroneus brevis or 
 tibialis posticus, by the joint action of which the 
 sole of the foot is kept in the normal middle 
 position. 
 
 Similarly when ataxy invades the arms, we see 
 dissociation of the synergic extension of the wrist 
 from the flexors of the fingers, so that the wrist 
 bends, and the grip is thereby rendered feeble and 
 ineffective. 
 
 When the patient stands up he has difficulty in 
 maintaining his balance, owing to the failure of 
 the centripetal impressions which are necessary to
 
 ON TABES DORSALIS 95 
 
 excite the muscular adjustments innervated by the 
 cerebellum. 
 
 Here the loss of conscious sensation when present 
 adds greatly to the difficulty ; for blunting of the 
 sensibility of the soles of the feet even by itself, as 
 has been proved experimentally by Vierordt and 
 Heyd, renders the maintenance of equilibrium very 
 unsteady. So long as vision is unimpeded, and 
 the vestibulo-cerebellar centripetal impressions 
 operative, standing is possible on a wide base ; it 
 is more difficult as the base of support is narrowed, 
 and is altogether impossible when the eyes are shut. 
 
 When the patient sits down, instead of doing so 
 gently with graduated flexion of his trunk, thighs 
 and legs, he bumps down on his seat, owing to 
 defective reflex antagonistic tone of his quadriceps 
 extensor muscles. 
 
 When he endeavours to maintain his equilibrium, 
 and also to walk, he assumes attitudes, and makes 
 compensatory movements which are almost patho- 
 gnomonic. Thus, if alone and unsupported, he 
 will walk with a constrained gait, his legs over- 
 extended and knees recurved (so as to prevent the 
 sudden giving way of the legs so well described by 
 Buzzard), his body bent forward, and with his eyes 
 steadfastly watching his feet. 
 
 But his actions vary with the conditions under 
 which he is placed. For the same patient, if 
 supported on both sides, and thus saved from the 
 necessity of attending to his equilibrium, will throw 
 out his leo's and dorsiflex his feet with excessive
 
 96 THE LUMLEIAN LECTURES 
 
 energy, and set them down aoain with equal 
 brusqueness, stamping his heels on the floor. 
 
 Throughout all the same purpose runs. It is the 
 attempt, clumsy at best, by conscious effort to 
 make up for the failure of a self-adjusting mechanism 
 which, under normal conditions, works better with- 
 out conscious interference. Clumsy though it be, 
 however, the best treatment that has been hitherto 
 devised for the palliation of ataxy is the assiduous 
 cultivation and judicious direction of these conscious 
 efforts at compensation in accordance with the rules 
 laid down by Frenkel and Forster, 
 
 The Tabetic Pupil. 
 
 One of the most important signs of tabes is the 
 condition of the pupil, first described by Argyll- 
 Robertson in 1869, and known universally after his 
 name. It is scarcely necessary to say that this 
 consists in the loss of the refiex contraction of the 
 pupils to light, while that on convergence and 
 accommodation remains. The frequency of the 
 Argyll- Robertson pupil in tabes was first noted by 
 Erb, as well as the usually associated loss of the 
 reflex dilatation to sensory or psycho-sensory 
 stimulation. 
 
 Reflex iridoplegia is not always present, nor 
 is it always absolute, there being frequently only 
 sluggish reaction, instead of total absence of con- 
 traction to light. But it is so frequent — in general 
 about three-quarters of the cases — and so character- 
 istic that the other symptoms of tabes must, in its
 
 ON TABES DORSALIS 97 
 
 absence, be well marked to allow of a definite 
 diagnosis of tabes to be made. 
 
 Sometimes it is unilateral, or more marked in 
 one eye than the other. It is, in a large proportion 
 of cases, associated with myosis, so that a peculiar 
 appearance is given to the eye of almost a patho- 
 gnomonic character. 
 
 But myosis is by no means universal ; sometimes 
 the pupils are of medium size, or even larger than 
 normal, and they may be unequal. The essential 
 and characteristic feature of tabes is the reflex irido- 
 plegia. This rarely passes into absolute iridoplegia. 
 This, however, sometimes occurs unilaterally, but 
 it is then more probably due to a local syphilitic 
 affection. Absolute iridoplegia, however, occasion- 
 ally gives way to reflex iridoplegia. 
 
 The outline of the pupil, also, generally under- 
 goes change, and this not infrequently precedes the 
 loss of light reflex. The tabetic pupil is rarely 
 quite circular, and it is often eccentric in position. 
 
 What has been said of the pupils of tabes is true 
 also of those of general paralysis, but here they are 
 more often unequal in size. Reflex iridoplegia 
 once complete is incurable ; but occasionally when 
 incomplete the reaction may return. It may exist 
 for many years (Erb) without any other obvious 
 symptoms either of tabes or general paralysis. 
 Usually it is associated with lymphocytosis of the 
 cerebro- spinal fluid, but, as I have found, this is 
 not invariable. 
 
 But one may say with Bumke that a permanent 
 7
 
 98 THE LUMLEIAN LECTURES 
 
 isolated reflex iridopleo^ia occurs only in tabes, 
 general paralysis, and as a consequence of con- 
 genital or acquired syphilis ; while the cases in 
 which, as a residuum of absolute iridoplegia, a 
 condition remains which is not distinguishable 
 from the Argyll- Robertson pupil, are so rare that 
 they may be practically neglected. The general 
 result, therefore, if we consider the syphilitic origin 
 of tabes and general paralysis as proven, is that 
 reflex iridoplegia is always a result of syphilis. 
 
 The pathology of the Argyll- Robertson pupil is 
 a subject of great difficulty, for we are still far from 
 possessing a complete knowledge of the mechanism 
 of pupillary variations. The pupil varies in size 
 not merely under conditions of illumination and 
 accommodation, but with the intraocular tension 
 and fulness of the vessels of the iris ; it contracts 
 with the orbicularis oculi, dilates under the influence 
 of sensation and emotion, and varies with psychical 
 activity in a manner extremely difficult to analyse. 
 We do not yet know for certain whether the retinal 
 elements which subserve light sensation are the 
 same as those which excite the reflex pupillary 
 contraction. There are conditions of the retina in 
 which vision may be afl^ected and the light reflex 
 continue, and vice versii : and Schirmer concludes 
 from pathological conditions that the parareticular 
 or amakrine cells of the inner granular layer are 
 the special centripetal organs of the light reflex. 
 Whether this is well founded or not is a question, 
 but both anatomical and clinical facts point to a
 
 ON TABES DORSALIS 99 
 
 differentiation in the optic nerve between the fibres 
 conveying light impressions and those which 
 excite pupillary reflex. In man, as in all animals 
 with semidecussation of the optic nerves, the 
 centripetal reflex pupillary fibres undergo also 
 partial decussation. Hence lesion of the optic tract, 
 as in Wernicke's well-known hemiopic test, not only 
 causes contralateral homonymous hemianopsy, but 
 also loss of the light reflex from the corresponding 
 retinal halves. 
 
 In front of or in the immediate neig-hbourhood 
 of the corpus geniculatum externum, the centripetal 
 pupillary tracts separate from those which form 
 the optic radiations, and pursue their course 
 towards the oculo-motor centres. But what this 
 course precisely is, is still a vexed question. 
 
 That the pupillary and other oculo-motor reflex 
 centripetal paths do not pass through the vault 
 or tectum of the corpora quadrigemina, is evident 
 from the results of experimental lesions of these 
 ganglia in monkeys, as described by Aldren Turner 
 and myself After such lesions the reaction of the 
 pupils was intact ; and many cases have been 
 related of tumours invading the corpora quadri- 
 gemina without affecting the pupillary reflex. 
 
 It has been held by some that the fibres of 
 Meynert's fascicnlns retro fiexiis (fig. ^il^'' which 
 courses, internal to the red nucleus, from the 
 oano-lion habenula^ towards the base, is the centri- 
 petal reflex pupillary path, and that therefore a 
 lesion of this tract would cause reflex iridoplegia.
 
 lOO 
 
 THE LUMLKIAX LECTURES 
 
 But no such lesion has ever been demonstrated 
 in the reflex 1 iridoplegia of tabes ; and von 
 Monakow has observed destruction of Meynert's 
 fasciculus retroflexus without any affection of the 
 
 Cospm 
 
 Fig. yi- — Frontal section through lenticular nucleus, showing 
 the fasciculus 7-eiroflcxiis. Ceni = corpus mamillare externum ; 
 CH = central tegmental tract ; chy = commissura hypothalamica ; 
 Ci = capsula interna ; Cm = corpus mamillare ; Cospm = commis- 
 sura supra-mamillaris; Csth = corpus sub-thalamicum ; Frtf = 
 fasciculus retroflexus ; ghb = ganglion habenulas ; Gi — latticed 
 layer of thalamus; Gp = globus pallidus ; Hi, H2 = Forel's 
 bundle from the red nucleus ; Lme = lamina meduUaris externa ; 
 Narc = nucleus arcuatus thalami ; Nc = nucleus caudatus ; Nl =^ 
 nucleus lateralis thalami ; NL = nucleus of Luys ; Ntg = red nucleus 
 of tegmentum ; Nvl = nucleus ventralis lateralis thalami; Pu = puta- 
 men ; Pp = Pes pedunculi : St = Stria terminalis ; Stz = Stratum 
 zonale thalami ; V3 = third ventricle ; Zi = zona incerta ; II = optic 
 tract (Obersteiner). 
 
 pupillary reflex. Moreover, after extirpation of 
 both eyeballs this tract remains free from de- 
 ofeneration. 
 
 The probability is that there is no direct con- 
 nection between the centripetal optic paths and
 
 ON TABES DORSALIS 
 
 lOI 
 
 the oculo-motor centres, but that the connection is 
 indirect, through intercalary or association cells, 
 which again form synapses with thQ oculo-motor 
 nuclei. But these and their respectiivie relations 
 to the centripetal reflex paths from euch eye are 
 matters more of hypothesis than of actual demon- 
 stration. 
 
 Fig. 38. — Diagram of relation of the optic tracts to the oculo-motor 
 centres (Bumke). a, b, or ai, bi, position of lesion causing unilateral 
 reflex iridoplegia, together causing double reflex iridoplegia ; c, 
 optic chiasma ; d, position of lesion causing Wernicke's hemiopic re- 
 action ; e, optic nerve ; R.A, L.A, right and left eyeball respectively. 
 
 In order to account for the various disorders 
 that may be observed in reflex pupillary innerva- 
 tion the accompanying diagram (fig. 38) on the 
 screen has been suggested by Bumke. This shows 
 how hemiopic, as well as unilateral or double reflex 
 iridoplegia, may occur according to the position ot 
 the lesion. 
 
 But it has been said, the only verified basis 
 of this scheme is Wernicke's hemiopic pupillary 
 reaction from lesion of the optic tract. No lesion
 
 I02 THE LUMLEIAN LECTURES 
 
 in the central grey matter of the aqueduct of 
 Sylvius has been established in connection with 
 the tabetic pupil, though Moeli has reported a 
 case of tumour compressing the hinder wall of 
 the third ventricle in which reflex iridoplegia was 
 said to have been present. 
 
 Whether in the conglomeration of nerve cells 
 which constitute the nucleus of the third nerve 
 there is an isolated group presiding over the 
 movements of the iris, or whether the iris is 
 represented more or less diffusely along with the 
 various muscles with w-hich it acts in association, 
 is not as yet positively determined. The experi- 
 ments of Henson and Volkers, however, are in 
 favour of the centres of the iris and ciliary muscle 
 being in the anterior part of the oculo-motor 
 nucleus, and their nerves in the most anterior 
 roots of the third nerve. 
 
 Bernheimer maintains, mainly on the results of 
 experiments on monkeys, that the centres for the 
 sphincters of the pupils are situated in the paired 
 small-celled nucleus termed the Edinger-Westphal 
 nucleus (fig. 39). 
 
 He states that the cells of the corresponding- 
 nucleus show degenerative changes after enuclea- 
 tion of the eye -ball ; that unilateral destruction 
 causes reflex iridoplegia in the corresponding eye ; 
 and that electrical irritation of this nucleus causes 
 contraction of the pupil of the same side. Bern- 
 heimer's experiments have been subjected to 
 searching and destructive criticism by Bach, and
 
 ON TABES DORSALIS 
 
 lo- 
 
 they certainly appear to me far from convincing". 
 They are opposed to many recorded and carefully 
 investigated clinical cases. Thus von Monakow 
 has observed cases of total ophthalmoplegia, 
 external and internal, in which there was com- 
 plete degeneration of all the cell-groups of the 
 oculo-motor nerve, with the exception of the 
 
 ^VS^i?'?"^*! 
 
 
 'mm 
 
 Si*' 
 
 ♦«i\\ 
 
 Fig. 39. — Frontal section throiijjh oculo-motor nuclei, i, Small- 
 celled paired Edinger-Westphal nucleus (after Bernheimer). 
 
 Edinger-Westphal nucleus. And conversely, the 
 Edinger-Westphal nucleus has been found de- 
 stroyed without impairment of the pupillary reac- 
 tion (v. Rad's case quoted by v. Monakow). As 
 against the localisation of a definite circumscribed 
 centre for the sphincter pupilhe, is also the fact 
 that the action of the pupil is usually unimpaired 
 in ophthalmoplegia dependent on acute lesion of 
 the oculo-motor nuclei {polioencephalitis superior).
 
 I04 THE LUMLEIAN LECTURES 
 
 A Strictly limited lesion of one particular cell- 
 group of the oculo- motor nucleus has never been 
 demonstrated in man in relation with reflex irido- 
 plegia or ophthalmoplegia interna. Marina, who 
 has investigated many cases of reflex iridoplegia in 
 tabes, has uniformly found the Edinger-Westphal 
 nucleus normal. 
 
 The pupillary fibres of the third nerve pass 
 to the sphincter pupilla", which is an involuntary 
 muscle, not directly but by the short root of the 
 ciliary ganglion, with the cells of which it forms 
 synapses, and thence by the short ciliary nerves 
 to the sphincter. 
 
 In reflex iridoplegia no lesion has ever been 
 demonstrated in the trunk of the third nerve. 
 Complete transverse lesion of this nerve, in addi- 
 tion to the paralysis of the external muscles which 
 it innervates, causes dilatation and complete immo- 
 bility of the pupil both to light and convergence — 
 absolute iridoplegia. It is, however, to be observed 
 that a less complete lesion of the third nerve may 
 produce external ophthalmoplegia without affecting 
 the pupil. Many such cases are on record. Gush- 
 ing has reported three cases of accidental injury to 
 the third nerve during the operation for excision 
 of the Gasserian ganglion, in which there was 
 temporary total paralysis of the external ocular 
 muscles, while the pupil remained unaffected, 
 reacting normally to light and accommodation. 
 
 From this brief review of our knowledge respect- 
 ing the reflex pupillary arc, it is evident that there
 
 ON TABES DORSALIS IO5 
 
 is Still great obscurity as to its precise constitu- 
 tion, and that though, theoretically, reflex irido- 
 plegia may be explained by interruption of certain 
 hypothetical centripetal paths of the light reflex, 
 no actual pathological changes have as yet been 
 demonstrated in the reflex iridoplegia of tabes. 
 
 The frequent association of reflex iridoplegia 
 with degeneration of the cervical region of the 
 cord, has led several observers — notably Rieger and 
 Forster — to place the centre for the light reflex 
 in this reoion. But it is obvious that in tabes 
 the loss of the light reflex and degeneration of 
 the posterior columns need not be causally related 
 to each other. And the medulla has been found 
 normal in two cases of general paralysis with irido- 
 plegia (Marina), and in a case of congenital Argyll- 
 Robertson pupils by Reichart. 
 
 Degenerations of the posterior columns, other 
 than tabetic, have also been found by various 
 observers with normal reaction of the pupils. 1 his 
 is notably the case in Friedreich's disease, in which 
 there is no abnormality (jf the pupillary reactions. 
 
 Bach, however, who at first was inclined to 
 place the centre of the light reflex in the cervical 
 cord just below the medulla oblongata, in his more 
 recent work relates various experiments made in 
 conjunction with Meyer, which lead him to believe 
 that at the spinal end of the medulla oblongata, 
 near the respiratory centre, there is an inhibitory 
 centre of the light reflex, to which the reflex centre 
 of the corpora quadrigemina is subordinate ; and,
 
 I06 THE LUTNILEIAN LECTURES 
 
 in addition, in close proximity, another centre 
 inhibitory of the reflex dilatation of the pupil.* 
 
 That such a complicated mechanism — which 
 reminds one of the cycles and epicycles of the 
 Ptolemaic astronomy — is needed to explain the 
 facts, is in the highest degree improbable, and that 
 the reflex iridoplegia of tabes, often of years' 
 duration, can be explained by persistent irritation 
 of these hypothetical inhibitory centres of pupillary 
 action has little in its favour. 
 
 Of all the investigations on the pathology of the 
 
 "- The facts on which Bach bases his hypothesis are as 
 follows : (i) Total division of the cord several millimetres 
 caudally from the fourth ventricle causes no change in the 
 pupillary reaction. At the moment of section there is a 
 temporary dilatation of the pupils, after which the light reflex 
 can be elicited as before. (2) Bilateral section of the medulla 
 at the spinal end of the fourth ventricle at a definite level 
 causes immediate immobility of both pupils to light. (3) 
 Unilateral division of the medulla at the same level on the 
 right side produces not a right-sided but a left-sided reflex 
 iridoplegia. (4) Exposure of the medulla, with its mechanical 
 and other irritation, generally suffices to cause the disappear- 
 ance of the light reflex or absolute iridoplegia. Witli this 
 there is marked myosis and often inequality ot tlie pupils. 
 This fixity of the pupils, even after persisting from a quarter 
 of an hour to one hour, at once disappears, and is replaced 
 by prompt reaction of the pupils when a section is made 
 through the middle of the fourth ventricle or at higher levels. 
 This reaction lasted in one case for a whole hour. (5) Uni- 
 lateral division of the medulla at the middle of the fourth 
 ventricle and brainwards therefrom, renders the previously 
 sluggish or absent light reflex again evident. The persisting 
 inhibition can be counteracted by cocaine or by the action 
 of ether.
 
 ON TABES DOKSALIS IO7 
 
 tabetic pupil in recent years, those of Marina 
 appear to me the most suggestive. 
 
 Marina has investigated, by the most approved 
 modern histologrical methods, thirteen cases of 
 general paralysis with normal pupils, twenty-three 
 cases with Argyll- Robertson, five cases of tabes, 
 and twenty-nine other cases with normal or ab- 
 normal pupils. In all the cases of tabes and 
 general paralysis with the characteristic pupils he 
 
 Fig. 40.-— Ciliary ganglion — normal (Marina). 
 
 found signs of degeneration in the ciliary ganglia 
 and short ciliary nerves. If the pupils were normal 
 the ciliary ganglia were also free from degeneration. 
 In one case of unilateral Argyll- Robertson pupil, 
 the ciliary ganglion of the affected eye was pro- 
 foundly altered, while that of the other eye was 
 practically normal. The changes in the ciliary 
 oanolion in tabes are of a chronic character, and 
 consist in a gradual wasting of the nerve cells, with
 
 ■I08 THE LUMLKIAN LECTURES 
 
 chromatolysis and distortion of the nuclei (figs. 40, 
 41 and 42). 
 
 Degeneration of the short ciliary nerves was 
 constantly found, and apparently secondary to that 
 
 Fig. 41. — Ciliary ganylioii from a case uf Central Paralysis with 
 Ar^^yll-Robertson pupils, bliowini; cicformity of the cells (Marina. 
 
 Fig. 42. — Ciliary ganglion from a case of Tabes with Argyll- 
 Robertson pupils, showing profound chromatolysis (Marina). 
 
 of the cranoflion cells. The roots of the oanolion 
 were always found normal ; so also, with one excep- 
 tion, were the oculo-motor nerves, the oculo-motor
 
 ON TABES DORSALIS 
 
 109 
 
 nuclei, the Edinger - Westphal nucleus, that of 
 Darkschewitch, as well as the grey matter sur- 
 rounding the aqueduct of Sylvius. 
 
 Besides the ciliary ganglion, the Gasserian gan- 
 glion showed degenerative changes in three out of 
 
 Bulbil s sirusler 
 
 .\f rfc/i/s sufltrior ' Vt^ 
 
 I^ ciliorfs brevfs 
 
 ' Rada brertjt ^eu 
 motoncL 
 
 RamusR 
 
 Rajnuji Iff 
 Gafighitn Gassen - 
 
 N irifriitronhL'aJ^^^ 
 
 aV pthmouJaJj-s 
 
 N ct/zcuva lon.(^ 
 
 V opticus 
 PleXLU) carolicas 
 
 Art en a carotis 
 Trajytijx optuxLS 
 
 J\ ocuhimntonuji 
 
 y Irigemwiis 
 
 Flc. 43. — Anatomy of the ciliary gant(lion in the dog (Piltz). 
 
 sixteen cases of general paralysis, and in three out 
 of four of tabes. The superior cervical ganglion 
 also showed degenerative changes in two out of 
 four cases of tabes ; while in general paralysis one- 
 third only exhibited degeneration, the other two- 
 thirds being doubtful or normal. 
 
 The results obtained by Marina are evidently of
 
 no THE LUMLEIAN LECTURES 
 
 the Utmost importance in reference to the pathology 
 of the tabetic pupil. 
 
 The physiology of the ciliary ganglion is still 
 very imperfectly understood. That it is not an 
 independent centre for the pupillary reflex is evident 
 from the fact that the pupil is quite immobile when 
 the third nerve is completely severed. 
 
 Anderson holds that the cells of the ciliary gan- 
 glion are purely of the sympathetic type ; and that 
 most of them are of this character is also held by 
 Marina. But the ciliary ganglion has three roots — 
 the short or motor root from the third, the long or 
 sensory root from the fifth, and the sympathetic 
 root from the carotid plexus. The long ciliary 
 nerves which come from the fifth, and which supply 
 the dilator of the iris, do not go through the ciliary 
 ganglion (fig. 43). 
 
 Anderson has not been able to trace any degen 
 erated fibres in oculo- motor, fourth, fifth, or sixth 
 nerves after extirpation of the ciliary ganglion in 
 the cat. If, therefore, he says, cells comparable 
 with the cells of the posterior root ganglia are 
 present in the ciliary ganglion, they must belong 
 to a type of cell not yet demonstrated, and send 
 non-medullated processes to the central nervous 
 system. 
 
 On the other hand, Marina has observed deeen- 
 eration in some of the cells of the ciliary ganglion, 
 probably of a sensory type, after lesions of the 
 cornea. And Bumm finds that the number of cells 
 which remain intact after section of the short ciliary
 
 ON TABES DORSALIS I I I 
 
 nerves, is again reduced by extirpation of the 
 superior cervical ganglion. A considerable number, 
 however, still remain, some of which, at least, appear 
 to be connected with the fifth. The ciliary ganglion 
 would therefore, if these observations are correct, 
 appear to be a complex structure. That some of 
 the sympathetic fibres which supply the dilator of 
 the pupil pass into the ciliary ganglion by way of 
 the sympathetic root from the carotid plexus, instead 
 of all going by the long ciliary nerves as is com- 
 monly held, is, I think, probable. For though 
 division of the ophthalmic branch of the fifth, as 
 Turner and I have demonstrated in monkeys, 
 causes contraction of the pupil from paralysis of 
 the dilator, this effect is only of temporary duration, 
 lasting usually only a week or so. And in man, as I 
 have ascertained, after extirpation of the Gasserian 
 ganglion for inveterate trigeminal neuralgia, the 
 pupil does not remain permanently affected, but 
 contracts and dilates as under normal conditions. 
 This is very different from the state of the pupil 
 after division of the dilator fibres as they emerge 
 from the upper dorsal roots. Even years after this 
 lesion the pupil remains smaller, and the eyeball 
 sunk, as compared with the other. Nor does 
 cocaine cause dilatation, as in the normal pupil, 
 except in relatively very large doses. There must, 
 therefore, be some other path for the dilator fibres 
 of the iris than the long ciliary nerves, and it is 
 probable that they go through the ciliary ganglion 
 by the sympathetic root of the carotid plexus.
 
 112 THE LUMLEIAN LECTURES 
 
 The ciliary ganglion, as Langley and Anderson 
 have shown, is, like sympathetic ganglia in general, 
 capable of being paralysed by nicotine ; so that 
 impulses to the sphincter pupillse are blocked and 
 ineffective, while direct irritation of the short ciliary 
 nerves causes contraction. 
 
 iS\ 
 
 (f^^ 
 
 w 
 
 Papillae 
 
 dextra. 
 
 sinistra 
 
 b 
 Fig. 44. — Excentric position of the pupils in general paralysis (Piltz). 
 
 The question is whether the degeneration found 
 by Marina in the ciliary ganglia in tabes, taken in 
 conjunction with what we know of the physiology 
 of these structures, throws any light on the path- 
 ology of the tabetic pupil. In this relation the 
 researches of Piltz appear to me to be of great 
 significance. Piltz calls special attention to the 
 fact, to which I have previously alluded, that the 
 pupils in general paralysis and tabes commonly 
 exhibit irregularities of outline, being angular, 
 jagged, or otherwise misshapen, and not infre- 
 quently eccentricall)- situated, instead of having the 
 smooth, circular contour and central position of 
 health (fig. 44). 
 
 This condition often precedes the development
 
 ON TABES DORSALIS 
 
 1 I 
 
 of the Argyll- Robertson pupil, and Joffroy and 
 Schrameck maintain that every alteration in the 
 form of the pupils — conoenital anomalies and 
 synechiie excepted — even though the light rcllex 
 be retained, sionifies that we have to deal with 
 
 Fig. 45. — Varieties of irregularity of the pupils in difterent forms nf 
 organic nervous and mental diseases fPiltzi. 
 
 Fig. 46. — Varieties of irregularity of the pupil produced experi- 
 mentally by stimulation of different twigs of the long and short ciliary 
 nerves (Piltz). 
 
 general paralysis, tabes, or syphilis, and that the 
 semeiological value of irregularity of the pupil is 
 not less than that of the Argyll- Robertson phe- 
 nomenon itself, which it often precedes (fig. 45). 
 This is also the opinion of Babinski, Piltz, and 
 others. 
 
 After complete exposure of the ciliary ganglion
 
 114 THE LUMLEIAN LECTURES 
 
 and ciliary nerves in the dog, and irritation of 
 various twigs of the short and long- ciliary nerves, 
 Piltz produced irregularities in the outline of the 
 pupil and in its position, such as are indicated in 
 the accompanying figure (fig. 46). 
 
 The similarity of these appearances to those of 
 the pupils in tabes and general paralysis is obvious, 
 and seems to justify the conclusion that these abnor- 
 malities are indications of irritation, paresis, or 
 paralysis of different portions of the iris, due to 
 pathological changes in the long or short ciliary 
 nerves or their ganglia. It has also been shown 
 by Dupuy-Dutemps that in the Argyll- Robertson 
 pupil the inner margin of the iris, instead of having 
 the normal cushion-shaped outline on section, is 
 extremely attenuated. This atrophy of the iris 
 may be general, or in cases where the pupil is 
 irregular it may affect only certain segments of the 
 iris. This atrophv is undoubtedly due to affection 
 of the short ciliary nerves. 
 
 It is difficult to conceive how such irregularities 
 of the pupil could be produced by any central 
 lesion ; and seeing- that they so often precede the 
 reflex iridoplegia of tabes, the presumption is 
 strong that both phenomena have essentially the 
 same anatomical basis, namely, degeneration of the 
 ciliary ganglia and short ciliary nerves. 
 
 Piltz's experiments enable us to understand the 
 irregularities in the outline of the pupils, but how 
 can we explain the loss of the light reflex, with 
 the retention of the contraction or accommodation }
 
 ON TABES DOKSALIS I I 5 
 
 We can, I think, cite analogous facts. We know- 
 that different nerve fibres in the same trunk may 
 be differently affected by destructive or toxic 
 agencies. The extensor muscles are principally 
 affected in the peripheral neuritis of alcohol, lead 
 and arsenic ; and pressure on the recurrent laryngeal 
 first affects the abductor nerves of the larynx. 
 Langley and Anderson have shown that nicotine 
 paralyses the sphincter of the pupil and ciliary 
 muscle long before causing paralysis of the external 
 ocular muscles ; and we know that the toxin of 
 diphtheria may affect the ciliary muscle alone, 
 leaving the actions of the pupil and the external 
 ocular muscles intact. It has also been mentioned 
 that partial lesions of the trunk of the third nerve 
 may cause ophthalmoplegia externa without impair- 
 ing the action of the pupil. And total ophthalmo- 
 plegia interna due to syphilis may under antispecific 
 treatment give way to a condition merely of reflex 
 iridoplegia. 
 
 There is therefore nothing improbable in the 
 assumption that in tabes the ciliary ganglion and 
 ciliary nerves are so affected that, though they 
 cannot transmit the reflex impulse of light to the 
 sphincter pupilla;, they can readily allow the more 
 powerful stimulus associated with accommodation 
 to pass through. It is also probable that the 
 myosis, which is so commonly associated with the 
 reflex iridoplegia in the tabetic pupil, is due to 
 degenerative changes of an irritative character 
 which are going on in the sphincter.
 
 Il6 THE LUMLEIAN LECTURES 
 
 In a similar manner may be explained the absence 
 of the dilatation which normally results from sensory 
 or emotional stimuli. 
 
 The dilatation of the pupil which occurs under 
 such conditions is practically independent of the 
 cervical sympathetic, and may be observed, though 
 to a less extent, when this nerve is divided. This 
 was first noted by Vulpian, and has been verified 
 by Bumke as well as by myself in cases of lesion 
 of the dilator roots of the cervical sympathetic. 
 
 And in tabes the dilator of the pupil is not para- 
 lysed, for under the influence of cocaine the pupil 
 enlarges as in the normal state, and all the irregu- 
 larities of its contour become accentuated. 
 
 Psycho-reflex dilatation is due, according to 
 Braunstein, to inhibition of the oculo-motor nucleus 
 acting through the third nerve, and therefore 
 through the ciliary ganglion, for it does not occur 
 in oculo-motor paralysis. Therefore, though the 
 exact mechanism may require further elucidation, 
 the probability is that the condition which blocks 
 the path of reflex pupillary constriction blocks also 
 that of psycho-reflex dilatation ; and that in the 
 degeneration of the ciliary ganglion we have the 
 explanation of the tabetic pupil. 
 
 If time had permitted, I might have offered some 
 considerations on the pathology of the disorders of 
 innervation of the bladder, visceral crises, arthro- 
 pathies, muscular atrophies, and other notable 
 symptoms of tabes ; but I found, as I filled page 
 after page on the questions to which i have devoted
 
 ON TABES DORSALIS I 1 7 
 
 these lectures, that it would be impossible for me to 
 deal effectively with them all, and it seemed to me 
 better to discuss a few points with some fulness of 
 detail, than range cursorily over the whole field. 
 Such as I have chosen I mio-ht have treated at 
 much greater length, but I shall be glad if I have 
 succeeded in showino- that in such a well-worn 
 subject as tabes there are still many problems of 
 surpassing interest, and worthy of further diligent 
 study and research.
 
 Il8 THE LUMLKIAN LECTURES 
 
 BIBLIOGRAPHY. 
 
 Anderson. Journal of Physiology, vol. xxxiii., No. 2, 1905. 
 
 Bach. Zeitschrift fiiv Angenheilknnde, Bd. xi., 1904. 
 
 Ballance and Purves Stewart. " The Healing of Nerves," 
 London, 1901. 
 
 Benedikt. "Tabes Fragen," 1901 ; International Medical Con- 
 gress, Moscow, 1897. 
 
 Bernheimer. v. Griife's Archiv.f. Ophthalmol., Bd. xliv. 
 
 Bethe. " Ueber die Regeneration peripheren Nerven," 
 Nenrolog. CentralbL, 1901, p. 720. 
 
 BiCKEL and Ewald. Pfliige/s Archiv, 1897, ^'^- ixvii. 
 
 Bourdon and Luys. Archives Generales des Medecine, 1861. 
 
 Brosius. Neurol. CentralbL, 1903, p. 606. 
 
 BuMKE. " Die Pupillen-Storungen bei Geistes und Nerven- 
 krankheiten," p. loi. 
 
 BuMM. Neurol. CentralbL, 1902, p. 423. 
 
 Campbell. " Localisation of Cerebral Function," 1905. 
 
 Collier and Buzzard. Brain, 1903, p. 559. 
 
 Gushing. Journal of the American Medical Association, April 8, 
 1905. 
 
 Dejerine et Thomas. " Maladies de la Moelle Epinieie," 
 igo2. 
 
 Dejerine et Thomas. " Semiologie du Systeme Nerveux." 
 
 Demeny. " Du role mecanique des Muscles Antagonistes 
 dans les Actes de Locomotion," Archiv. de Physiol., 1890. 
 
 Duchenne. " Die I'Alaxie Locomotrice 'Progressive," Archiv. 
 Gen. de Med., December, 1858, et. seq. 
 
 Duchenne. " De I'Electrisation Localisee," 1861. 
 
 DuPUY-DuTEMPS. Annalis d'Ocnlistique, September, 1905. 
 
 Edinger and Helbing. " Verhandlungen des XVI. Con- 
 gresses fiir innere Medicin," Wiesbaden, 1898. 
 
 Edinger. " Einiges liber Wesen und Behandlung der Tabes."
 
 ON TABES DORSALIS I 19 
 
 Edinger. " Verhandlungen des XVI. Congresses fur innere 
 
 Medecin," 1898. 
 Egger. " De la Sensibilite Osseuse," Journal de Physiologie, 
 
 1899. 
 Erb. " Syphilis und Tabes," Berlinev klin. Wochenschrift, 1904, 
 
 Nos. I to 4. 
 Erb. Die deutsche Kliitik, 1905 (Sonderabdrvick). 
 Eyre and Flashman. British Medical Journal, October 28, 
 
 1905. 
 Ferrier and Turner. " Experimental Lesions of the Cor- 
 pora Quadrigemina," Brain, 1901. 
 Flechsig. " Tabes u. System-Erkrankung.," Neurol. Centialhi, 
 
 1890. 
 FoRSTER. " Die Physiologie und Pathologic der Coordina- 
 tion," Jena, 1902. 
 FouRNiER. " Affections Parasyphilitiques." 
 FouRNiER et Raymond. " Paralysie generate et Syphilis," 
 
 Acad, de Med., 1905. 
 Friedlander. " iJber Storungen der Gelenksensibihtat bei 
 
 Tabes Dorsalis,'' Neurol. Centralhl., No. 13, 1905. 
 Gowers, Sir W. British Medical Journal, July 8, 1905. 
 Gull. " Cases of Paraplegia," Guy's Hospital Reports, 185S. 
 Head. " The Afferent Nervous System from a New Aspect," 
 
 Brain, 1905, p. 99. 
 HiRSCHL. " Die Aetiologie der progressiven Paralyse," Wien, 
 
 1896. 
 HiTziG. " Ueber trauniatische Tabes und die Pathogenese 
 
 der Tabes in allgemein," Neurol. Centralbl., September, 1894. 
 Holmes and Grainger Stewart. " Symptomatology of 
 
 Cerebellar Tumours," Brain, 1904, p. 522. 
 Homen. "Strang- und System-erkrankungen des Riicken- 
 
 marks," " Handbuch der path. Anat. des Nervensystems," 
 
 Abth. 3, Flatau, Jacobsohn und Minor, Berlin, 1903. 
 JoFFROY and Schrameck. Revue Neurol., 1902, p. 275. 
 Kennedy. *' On the Regeneration of the Nerves," Phil. 
 
 Trans. Roy. Soc, London, 1897, vol. clxxxviii., p. 257. 
 Koster. *' Zur Physiologie des Spinalganglien und der 
 
 trophischen Nerven, so^vie zur Pathogenese der Tabes 
 
 dorsalis," Leipzig, 1904. 
 Kron. Deutsche Zeitschr.f. N ervenheilh . , 1898. 
 Lesser. Berlin, klin. Wochenschr., 1904.
 
 I20 THE LUMLEIAN LECTURES 
 
 V. Leyden und Goldscheider. " Die Erkrankungen des 
 
 Riickenmarks und der Medulla oblongata," Wien, 1904. 
 V. Leydex. " Die graue Degeneration der hinteren Riicken- 
 
 marks-strange," 1863. 
 Marie.' " Maladies de la Moelle," 1S92. 
 Marie et Glullain. " Les Lesions du Systeme Lymphatique 
 
 Posterieur de la Moelle sont TOrigine des Processus 
 
 AnatomoPathologiques du Tabes," Revue Neurol., 1903, 
 
 No. 2. 
 Marina. Deutsche Zeitsckr. fur Neyvenheilhunde, 1901, I)d. xx. 
 Marina. Annali di Nevrologia, 1901, Fasc. iii. and iv. 
 Marinesco. " Recherches sur les Lesions des Cellules des 
 
 Ganglions Epinaux dans le Tabes," Revue Neurol., Tom. viii., 
 
 1900, p. 1,125. 
 Marinesco. " Les Troubles de la Pallesthesie et leur Co- 
 existence avec I'Anaesthesie \'ibratoire,''' Seinaine Medicale, 
 
 November 29, 1905. 
 MiLiAN. " La Nature du Tabes," La Syphilis, January, 1904. 
 Minor. Neurol CeutralbL, 1904, p. 146. 
 Moebius. " Ueber die Tabes," 1897. 
 V. MoNAKow. Gehirn. Pathol., Bd. ii., 1905, p. 1,033. 
 MoTT. Trans. Path. Soc., Loud., 1900, vol. li. 
 Mott. "Tabes in Asylum and Hospital Practice," Arcliiv. 
 
 of Neurol., vol. ii., 1903. 
 MoTT and Sherrington, Proc. Roy. Soc, 1895, "vol. Ivii. 
 MuNK. " Ueber den Folgen der Sensibilitatsverlustes der 
 
 Extremitat fiir dessen Mobilitat," Sitzungsherichte d. k. Acad. 
 
 d. Wissenschaften, Berlin, 1903. 
 Nageotte. " Pathogenic du Tabes dorsal," La Presse Medicale, 
 
 December 10, 1902, and January 3, 1903. 
 NoNNE. Fortschritt. d. Medicin, 1903. 
 NoNNE. " Syphilis und Nervensystem," 1902. 
 NoNNE. " Zur Casuistik der Betheiligung der peripheren 
 
 Nerven bei Tabes dorsalis," Archiv. f. Psychiatrie, Bd. xix. 
 Oppenheim und Siemerling. " Beitrage zur Pathologic der 
 
 Tabes dorsalis und der peripheren Nervenkrankung.," 
 
 Archiv. f. Psychiat. und Nervenkrank., Bd. xviii^, 1887. 
 Orr and Rows. " System Lesions of the Posterior Columns 
 
 in General Paralysis," Brain, 1904, p. 461. 
 Orr and Rows. " Lesions of Spinal Cord, the Result of 
 
 Absorption from Localised Septic Foci," Rev. of Neurol., 
 
 January, 1906.
 
 ON TAHES IJOKSALIS 12 1 
 
 Page May. "Segmental Representation of Movements in 
 the Lumbar Spinal Cord," Phil. Trans., vol. clxxxviii., I'^gy. 
 Philippe. These de Paris, 1897. 
 PiERRET. " Siir la Sclerose des Cordons Posterieurs," Archiv. 
 
 de Physiol., 1871, Th. iv., p. 364. 
 Pii.TZ. Neurol. Centralb., 1903, p. 662, et. seq. 
 PuRVEs Stewart. Brain, 1901, p. 229. 
 PuRVES Stewart. Edin. Med. Journ., 1906, p. 429. 
 Redlich and Obersteiner. " Ueber Wesen der Pathogenese 
 der tabischen Hinterstrangdegenerationen," Vorlnufige Mit- 
 theilnngen, Vienna, 1894. 
 Rydel and Seiffer. Archiv. f. Psychiatrie, 1903. 
 Richer. Physiologic Artisiiqne de r Homme en Mouvement, 1894. 
 Robertson, M'Rae and Shennan. Rev. of Neurol, and 
 
 Psychiat., 1903, vol. i., p. 225. 
 Robertson and M'Rae. Rev. of Neurol, and Psychiat., 1905, 
 
 vol. iii., p. 321. 
 Robertson. Morisonian Lectures, Rev. of Neurol., 1906, 
 
 vol. iv. 
 Robinson, Louis. " Darwinism in the Nursery," Nineteenth 
 
 Century, November, i8gi. 
 Romberg. " Lehrbuch der Nervenkrankheilen des Mens- 
 
 chen," ist Ed., 1840; 3rd, 1857. 
 ScHUPPEL. Archiv. ftir Heilhtnde, 1874, Bd. xv. 
 Sherrington. Schdfer's Physiology, vol. ii., p. 843. 
 Spielmeyer. " Ein Beitrag zur Pathologic der Tabes," 
 Archiv. f. Psychiat. u. Nervcnkrankhtiten, 1905, Bd. xl., p. 369. 
 Stroebe. "Ueber Veianderungen der Spinalganglien bei 
 Tabes dorsalis," Centralbl. f. altgem. Pathol, u. pathalog. Anat., 
 Bd. v., 1904. 
 Thomas et Hausek. " Etudes sur les lesions radiculaires et 
 ganglionaires du Tabes," Nouvelle Iconographie de la .Sal- 
 pet ric re, 1902. 
 Thomas. Societe de Neurol., Paris, March, 1904; ref. in 
 
 Neurol. Centralbl., 1905, p. 495. 
 Trepinski. " Die embryonale Faser-Systeme in der Hin- 
 
 terstrangen, Archiv. f. Psychiat., 1898. 
 VuLPiAN. Maladies de la Moi'lle, 1879. 
 Warrington. "The Cells of the Spinal Ganglia," Brain, 
 
 1904, p. 297. 
 Weigert, C. Neurol. Centralbl., 1904, p. 73. 
 
 9
 
 122 THE LUMLEIAN LKCTUKES 
 
 WiDAL, SicARD, et Ravaut. " Cytodiagnose du Tabes," 
 
 Revue Neurol., 1903, p. 289. 
 Williamson. British Medical /oiiiiial, January 16, 1904; Med. 
 
 Chronicle, May, 1905. 
 WoLLENBERG. " Untersuchungen iiber das \'erhalten der 
 
 Spinalganglien bei der Tabes dorsalis." Archiv. f. Psychiat. 
 
 n. Nervenkrank., Bd. xxiv., 1902.
 
 UNIVERSITY OF CALIFORNIA LIBRARY 
 
 Los Angeles 
 
 tfi'nt^pok is DUE on the last date stamped below. 
 
 12^ 
 
 M- 
 
 y 
 
 OCT 8 199^