The history of neuromyelitis optica. Part 2: ‘Spinal amaurosis’, or how it all began

Neuromyelitis optica (NMO) was long considered a clinical variant of multiple sclerosis (MS). However, the discovery of a novel and pathogenic anti-astrocytic serum autoantibody targeting aquaporin-4 (termed NMO-IgG or AQP4-Ab), the most abundant water channel protein in the central nervous system, led to the recognition of NMO as a distinct disease entity in its own right and generated strong and persisting interest in the condition. NMO is now studied as a prototypic autoimmune disorder, which differs from MS in terms of immunopathogenesis, clinicoradiological presentation, optimum treatment, and prognosis. While the history of classic MS has been extensively studied, relatively little is known about the history of NMO. In Part 1 of this series we focused on the late 19th century, when the term ‘neuromyelitis optica’ was first coined, traced the term’s origins and followed its meandering evolution throughout the 20th and into the 21st century. Here, in Part 2, we demonstrate that the peculiar concurrence of acute optic nerve and spinal cord affliction characteristic for NMO caught the attention of physicians much earlier than previously thought by re-presenting a number of very early cases of possible NMO that date back to the late 18th and early 19th century. In addition, we comprehensively discuss the pioneering concept of ‘spinal amaurosis’, which was introduced into the medical literature by ophthalmologists in the first half of the 19th century.


Background
Neuromyelitis optica (NMO, Devic's syndrome) is a rare syndrome characterised by optic neuritis and myelitis [1]. For many decades, NMO was considered a clinical subtype of multiple sclerosis (MS). However, the discovery of pathogenic autoantibodies to aquaporin-4 (AQP4) in patients with NMO, which are absent in MS, led to the recognition of NMO as a disease entity in its own right [2][3][4] and generated strong and persisting interest in the syndrome. While the history of MS has been extensively studied, relatively little is known about the early history of NMO [5].
In Part 1 of this series of articles [5], we focused on the late 19th century, when the term 'neuromyelitis optica' was first coined. We traced the origins of this peculiar term in the 19th century French-, English-, and German-language literature and followed its definition's meandering evolution throughout the 20th and into the 21st century.
Here, in Part 2, we turn the spotlight onto the very beginnings of the recognition of NMO as a distinct syndrome and re-present and discuss a number of early 19th century reports on cases of possible NMO, all of which precede Eugéne Devic (1858Devic ( -1930 and Fernand Gault's (1873Gault's ( -1936 disease-defining work on NMO, which appeared in 1894, by years or even decades, among them a detailed clinical case description and an early post-mortem analysis. These early reports demonstrate that the rare and intriguing coincidence of acute spinal cord and optic nerve damage characteristic for NMO had caught the attention of the medical world much earlier than previously thought. At that time, the syndrome still went under another name: that of 'amaurose spinal' (or 'spinal amaurosis' in English and 'Spinale Amaurose' in German). That term had been introduced by French ophthalmologists in the early 19th century to refer to the unusual coincidence of spinal cord and optic nerve disease. The term remained in use throughout the entire 19th century but fell into oblivion after the turn of the century, when 'neuromyelitis optica', the designation proposed by Devic and Gault in 1894, increasingly prevailed. With that change in nomenclature, awareness of the long history of NMO as a subject of study in ophthalmology declined rapidly, and most of the early observations of NMO reported as 'spinal amaurosis' re-presented here have been almost completely forgotten.

Emergence of a novel concept
Julius Sichel and the concept of 'amaurose spinal' (1832/ 1837) The years following the French revolution saw a dramatic decline in the field of ophthalmology in France. In 1846, Phillip Franz von Walther (1782-1849), one of the founding fathers of scientific surgery and ophthalmology in Germany, a pupil of Georg Joseph Beer (1763-1821) and co-editor (together with Albrecht von Graefe ) of the Journal für Chirurgie und Augenheilkunde, could write: "Unbegreiflich ist, wie in einem Lande, in welchem früher die ersten Anfänge einer rationellen und wissenschaftlichen Begründung der Augenheilkunde durch Maître-Jan, St. Yves, Méjan, Daviel, J.L. Petit u. a. sich zeigten, mit dem Eintritt der Revolution der Faden der Entwicklung so ganz abreißen und für mehrere Dezennien totale Verdunklung entstehen konnte" [It is beyond comprehension how, in a country in which earlier the beginnings of a rationale and scientific foundation of ophthalmology were laid by Maître-Jan, St. Yves, Méjan, Daviel, J. L. Petit and others, the thread of development could break so completely and total obfuscation emerge] [6]. Von Walther's harsh verdict was shared by many outside France, as carefully documented by Julius Hirschberg   [7]. When ophthalmology was revived in France in the 1830s and 1840s, the renaissance was mainly driven by a number of migrants, among whom Julius Sichel (1802-1886) must be considered the most important figure. Sichel (Fig. 1) was a pupil of Johann Lukas Schönlein (1793-1864)one of the main modernisers of medicine in Germany (today best remembered for coining the terms 'tuberculosis' and 'haemophilia', as well as for the description of Schönlein-Henoch's purpura) and teacher of Rudolf Virchow (1821-1902)and of Christoph Friedrich Jäger (1784-1871), a former assistant to Beer. Sichel went to Paris in 1830 and founded one of the first French ophthalmological policlinics ('dispensaire') there in 1833. In 1834, he was made a French citizen. In 1837, he published his Traité de l'ophthalmie, la cataracte et l'amaurose [8], which soon became one of the most influential textbooks in its field.
It was in that work that Sichel introduced the concept of 'amaurose spinale' [spinal amaurosis], proposing that amaurosis may be linked to spinal cord disease in some cases: "L'amaurose peut également être symptomatique d'une affection de la moelle épinière ( …) Mais ce qui constitue le caractère principal de cette espèce d'amaurose, c'est là concomitance de symptômes qui se rapportent particulièrement à une affection co-existante de l'axe spinal ou d'une partie du cordon de la moelle épinière" [Amaurosis can also be a symptom of a disease of the spinal cord (...) What constitutes the main character of this type of amaurosis is the concomitant presence of symptoms that particularly indicate co-existing affection of the spinal axis or of a part of the spinal cord].
Sichel dedicated a whole chapter (entitled "Genre VI. -Amaurose spinale") to the topic, in which he points to the importance of proper neurological examination in patients presenting with amaurosis: "Ces détails sur l'exploration de la moelle épinière sont d'autant plus importants, que de l'exactitude avec laquelle l'examen de la colonne est fait dépend la certitude du diagnostic, et la possibilité de confirmer par de nouveaux faits l'existence d'une connexion étroite entre les affections de la moelle épinière et certaines espèces d'amaurose" [The details of the exploration of the spinal cord are all the more important as the accuracy of the spinal examination determines the certainty of the diagnosis and the possibility of confirming by new findings the existence of a close connection between the affections of the spinal cord and of certain types of amaurosis] [8]. Showing profound understanding of spinal cord function, he recommended physicians should pay attention in such patients not only to paralysis and numbness of the upper and lower limbs, but also to symptoms such as respiratory distress, pain, girdle-like dysesthesia ("engourdissements"), jerks and involuntary contractions of the extremities, convulsions, arrhythmia, circulatory problems, incontinence, urinary retention or constipation, impotence, and amenorrhoea. He even pointed to the potential occurrence of brainstem symptoms such as dysphagia and oculomotor disturbances: "Quelquefois il s'y joint des symptômes d'un dérangement dans la motilité du globe oculaire, diplopie" [Sometimes there are symptoms indicating disturbance in motility of the globe of the eye, diplopia]. Brainstem involvement has been recently rediscovered as a frequent feature of both AQP4-immunoglobulin G (IgG) and myelin oligodendrocyte glycoprotein (MOG)-IgG-positive NMO [1,9,10], and respiratory insufficiency due to brainstem or upper cervical cord lesions is nowadays the most common cause of death from NMO [11]. Similarly, tonic brainstem attacks have been shown to occur frequently in AQP4-IgG-positive NMO, and seizures have been observed both in AQP4-IgG-positive and in MOG-IgG-positive NMO in recent studies [12][13][14][15][16][17][18][19].
Regarding the question of how spinal cord disease and amaurosis could be related, Sichel could only speculate.
He referred to the network-like organisation of the nervous system and posited a direct link between the eye and the spinal cord: "La continuité de toutes les parties du système nerveux, et les rapports directs de liaison entre l'appareil nerveux de l'organe de la vision et l'axe spinal, sont là pour expliquer l'origine des amauroses de cette espèce" [The continuity of all parts of the nervous system, and the direct connections between the nervous apparatus of the organ of vision and the spinal axis, explain the origin of this sort of amaurosis].
In accordance with contemporary ideas on disease pathogenesis, he distinguished irritative ("congestive ou éréthique nerveuse") and torpid cases of spinal amaurosis. Among the torpid cases he counted, inter alia, those occurring during excessive breast feeding "bei nervösen und reizbaren Frauen" [in nervous and irritable women]. Interestingly, recent studies indeed suggest an increase in disease activity in NMO during the first year post partum [20,21]). Among the irritative cases he counted, besides cervical injuries and many others, cases caused by toxins, but frankly conceded: "Nous ne possédons que fort peu de données sur cette variété de l'amaurose spinale" [We have very little in the way of data on this variety of spinal amaurosis]. However, he strongly believed that besides these conditions, which he classified as 'functional', also 'organic' diseases of the spinal cord, which acted by "inflammation ou congestion", could cause amaurosis and recommended "recueillir avec un soin scrupuleux tous les matériaux propres à jeter du jour sur ce sujet" [to collect with scrupulous care all the proper materials to throw light on this subject].
Whether Sichel personally attended a patient with true NMO is unclear. In his Traité de l'amaurose ou de la goutte-sereine, Charles Deval (1806-1862), an ophthalmologist and former student of Sichel, later recollected the case of a young woman with partial visual loss, in whom Sichel (in Deval's presence) diagnosed "sub-inflammation de la moelle épinière" [subinflammation of the spinal cord] and whom he treated with leeches [22]. In his own writings, Sichel makes reference to a 14-year-old young girl whom he had seen in 1832. This girl presented with bilateral visual deficiency with (transiently) dilated pupils associated with spinal pain and contractions of the extremities. Complete remission was achieved following "un traitement révulsif, joint aux antispasmodiques, à la valériane, etc." [revulsive treatment, combined with antispasmodics, valerian, etc.] [8]. Whether the two reports relate to the same patient remains unknown.
Sichel was well aware of the obvious limitations of his recommendations, which he could not found on the authority of previous authors, but was convinced about the potential clinical relevance of his observations: "Nous sentons vivement le besoin de nouveaux éclaircissements sur ce point obscur de la pathologie oculaire que nous osons seulement effleurer dans l'intention d'y appeler l'attention des praticiens" [We feel strongly the need for further clarification on this obscure point of ocular pathology that we only dare to touch with the intention of calling it to the attention of practitioners] [8]. This call would be answered sooner than Sichel may have expected.
Carron du Villards (1838): "Myelitis and complete amaurosis" Besides Sichel, it was Charles Joseph Frédéric Carron du Villards (1799?-1860) (Fig. 2), who would cast some light into the 'darkness' described by von Walther. Carron du Villards studied with Antonio Scarpa (1752-1832)the famous Pavian surgeon and anatomist, first physician to Napoleon I, pupil of Giovanni Battista Morgagni (1682-1771), one of the founding fathers of pathological anatomy, author of the first ophthalmological textbook in the Italian language [23], and once characterised as "the highest culmination of the Galenic tradition of ophthalmology" [24] and then moved to Paris in 1828, where he took part in Jacques Lisfranc's (1790-1847) operation courses and later, together with Salvatore Furnari (1808-1866), founded his own ophthalmological dispensaire [7]. Being originally a native of Savoy, he was naturalised in 1832 and would later describe himself as an "français d'adoption" [25] (which did not restrain him from sharing von Walther's pessimistic perspective of contemporary French ophthalmology, complaining in 1838: "Pourquoi depuis 1799, la France a-t-elle cessé de payer son tribut à l'edifice de la science ophthalmologique?" [Why did France after 1799 stop paying its tribute to the construction of scientific ophthalmology?] [26]). However, he never settled in one place for long. In 1888, 28 years after his death, an obituary appeared in the Annales d'Oculistique [27], which, along with Hirschberg's notes, allows us to retrace the many stations of his restless and adventurous life: 1841 found him in Amsterdam, 1843 in Luxembourg, and his subsequent travels took him to Norway, Greece, Sphacteria, Candia, Tripoli, Tangiers, Liberia, Sierra Leone, Havana (where he worked for 2 years and survived two invasions), Puerto Rico, Mexico, Chile, Peru and Venezuela [7,28,29]. In Mexico, he bore the title of a general in the medical service; three times he was a castaway, and once he was shot and wounded [7]. In 1858, Carron du Villards arrived in Rio de Janeiro, whereunder imperial protectionhe would found the first public ophthalmological hospital, and where he was to die in 1860.
Carron du Villards was a prolific writer, whose Guide pratique pour l'etude et le traitement des maladies des yeux (Brussels, 1838) [25] is considered one of the most important ophthalmological textbooks of the period [7,30].
It is in this textbook that Carron du Villards speaks about "[une] espèce d'amaurose ( …) qui est sympathique d'une affection aiguë ou chronique de la moelle épinière" [a type of amaurosis (…) that is sympathetic to an acute or chronic affection of the spinal cord] and, importantly, declares, "j'ai vu plusieurs individus atteints de myélite qui furent complètement amaurotiques, pendant toute la durée de la maladie, la cécité disparut en même temps que l'inflammation intra-vertébrale" [I saw several individuals with myelitis who were completely amaurotic throughout the duration of the disease; the blindness disappeared at the same time as the intravertebral inflammation]. Like Sichel, he refers to the coincidence of optic nerve and spinal cord disease by using the term 'amaurose spinale'.
Based on his observations, he, just like Sichel before him, points to the need of carefully examining patients with amaurosis for signs of spinal cord inflammation, recommending, "il faut examiner avec soin la colonne vertébrale et promener sur tout son trajet des corps très-chauds ou très-froids, qui déterminent presque toujours une douleur ou tout au moins une sensation désagréable dans le point malade "[the vertebral column must be carefully examined, and very hot or very cold objects, which almost always generate pain -or at least an unpleasant sensationat the site of disease, must be moved all the way along the spine] [25].
Regarding therapy, he suggests bloodletting, which was considered to have "antiphlogistic" effects by some contemporary authors [31], and the application of leeches along the spine, in addition to other remedies of the time such as cupping, blistering and moxa. The former recommendation is interesting in the light of recent studies demonstrating that plasma exchange is effective in patients with acute NMO [32][33][34][35]. Given the enormous amounts of blood extracted by physicians performing phlebotomy in the 19th century (up to 2000 cc within the first 24-48 h, greatly overestimating the total blood volume, or until the patient fainted) [36], it is tantalising to speculate as to whether repeated bloodletting might indeed have had some therapeutic effect [31].
We carefully perused Scarpa's works on ophthalmological topics. However, it seems that Carron du Villards' teacher never made any mention of spinal cord diseases as a cause of amaurosis in his writings.

Pétrequin (1841): an early post-mortem report
Carron Du Villards' original works are numerous (a comprehensive list can be found in Hirschberg's Geschichte der Augenheilkunde [7], volume XIV, 3, §568). Among them is a short letter to a certain Joseph Pierre Eléonor(d) Pétrequin (1809-1876; see [37] for a biographical sketch) (Fig. 3) which appeared in the first volume of the Annales d'Oculistique et de Gynécologie [25] and marked the beginning of a debate about the best method for surgical treatment of cataract [25,[38][39][40]. It was the same Pétrequin, a then famous surgeon at the Hôtel-Dieu in Lyon with a profound interest in the history of medicine [41,42], who would shortly afterwards publish a further case of what Sichel and Carron du Villards had described as 'amaurose spinale'.
The patient, a 19-year-old male weaver ("lymphatique et scrophuleux quoique robuste" [lymphatic and scrofulous but robust]) from St.-Joux (Rhóne) experienced rapid-onset bilateral visual loss in May 1838, accompanied by continuous orbital pain. When the patient first presented on 3rd of June, Pétrequin found him completely blind with an amaurotic facies; the pupils were dilated and insensible to light. Pétrequin started to treat his patient with bloodletting and leeches, and, in addition, with purgatives (eau de Sedlitz, jalap, calomel, aloe) and a blister in the neck (as a counter-irritant). These measures were followed by an increase in vision and amelioration of the patient's headache. However, on 22nd of June the young man developed lumbar pain, dysaesthesia of the legs and urinary retention. Shortly afterwards, he developed severe paraparesis (plegia of the right leg; capable of elevating the left leg from the bed). At the same time, his sight continued to improve: by 4th of July he was able to count fingers and recognise bright colours, and by 11th of July, after seven frictions with a nux vomica tincture, vision had further increased to a degree that he was able to recognise individuals at eight steps' distance. On 18th of July, after continued treatment with nux vomica frictions, he was able to decipher large letters. By contrast, the paralysis continued and, 1 week later decubitus at the trochanters and the sacrum as well as urinary incontinence were noted, and the patient "s'affaiblit et dépérit peu à peu"; "le malade n'offre pas d'espoir" [weakens and decays little by little; the patient does not give reason to hope]. Over the next few days the decubital ulcers became worse ("disséquent tout le basin"), and the patient subsequently died on 2nd of September.
Besides respiratory insufficiency, sepsis from decubital ulcers and/or urinary retention was one of the most common causes of death from NMO well into the twentieth century, resulting in many patients not surviving their first relapse. Thisand the lack of long-term data in many other cases [44] may explain why NMO was long considered to be a mostly monophasic disease. Mortality has dramatically declined in recent years with routine use of high-dose steroid treatment, plasma exchange and immunosuppressants (in particular B celldepleting agents) [1,11], and with more and more patients surviving the first attack, NMO was found to be a relapsing disease in the vast majority of cases, especially in AQP4-IgG-seropositive patients [1].
What makes Pétrequin's report particularly valuable is the fact that it is accompanied by a detailed post- mortem report. While the pathologist found the brain, the cerebellum, and the spinal meninges unremarkable, he noted brownish-grey discoloration and softening of the spinal marrow beginning at the level of the 10th dorsal nerve, which he classified as the remnants of myelitis. The retina and optic nerves were considered healthy except for capillary congestion. The latter finding, although it may well reflect the restricted diagnostic means of the time (apparently, no microscopic analysis was performed), is in accordance with the patient's visual improvement as already appreciated by Pétrequin: "aucune lésion profonde ne formait obstacle à la guérison" [no profound lesion was present as an obstacle to healing]. Current studies have found complete recovery from optic neuritis (ON) in as many as 33% of attacks in NMO [1]. Pétrequin's report was made soon made known to an audience outside France by being re-printed in the second volume of the Annales de la Société de Sciences Naturelles, de Bruges [45].
Of note, two further brief reports on patients with paresis and visual disturbance can be found elsewhere in Pétrequin's writings [46]: A 26-year-old woman whose left eye had been "sujet à des ophthalmies" for a long time was admitted to the Hôtel-Dieu in Lyon in 1828 with paresis of all extremities and urinary retention requiring catheterisation; she recovered from paresis after 4 years. A 55-year-old woman developed transient amaurosis and hemiplegia in 1833 and subsequently completely recovered (following treatment with an alcoholic extract of nux vomica). Importantly, Pétrequin believed his patients' symptoms to have been caused by what he calledusing an originally Galenic termdyscrasia ("amaurose dyscrasique"), i.e. an abnormal, pathological composition of the blood. This is highly interesting in the light of the fact that most later authors would attribute ON in patients with spinal cord disease either to damage to the spinal origin of the sympathetic nerve (and, in consequence, nutritional damage to the optic nerve as a result of disturbed blood supply) or to ascending myelitisand thus to an anatomic or at least spatial connection between the two remote sites. It would take until 1896 for the two latter theories to be challenged by Karl Katz (1869Katz ( -1944, a doctoral student to Theodor Karl Gustav von Leber  in Heidelberg, who concluded: "die Gleichartigkeit der pathologisch-histologischen Veränderungen [weist] mit Notwendigkeit auf eine gemeinschaftliche im Blut circulierende Noxe hin "[the similarity of the pathohistological changes [indicates] necessarily a shared noxa circulating in the blood], and another century for two such 'noxae'pathogenic autoantibodies to AQP4, a water channel protein abundantly expressed in the central nervous system (CNS) [2,3,[47][48][49], and pathogenic autoantibodies to MOG [10,[50][51][52][53][54][55] to be identified in the serum of patients with NMO.
In pondering about some sort of dyscrasia underlying 'spinal amaurosis', Pétrequinknowingly or unknowingly possibly followed Konrad Johann Martin Langenbeck (1776-1851), who had stated already in 1818 in his Reflexionen über die Natur, Ursachen und Heilung des schwarzen Stars: "Gehen wir auf die Veranlassungen zurück, so finden wir in so vielen Fällen die Amaurosis offenbar als Folge mancherley Dyscrasien, deren erste Wirkung Entzündung ist" [regarding causation, we find that in many cases amaurosis is plainly the result of dyscrasias, the first effect of which is inflammation] andhimself drawing a parallel to humoral pathologydiscussed that "Ablagerungen irgendeines Krankheitsstoffes" [deposits of some sort of pathogen] may underlie at least some cases of amaurosis. It is fascinating to see that modern histopathology indeed revealed IgG and complement deposits (Ablagerungen) in optic nerve and spinal cord lesions from patients with NMO [56][57][58][59], a feature not present in most cases of classic MS [60,61]).
'Spinal amaurosis' in the medical literature of the first half of the 19th century Sichel's concept of 'spinal amaurosis' was soon further popularised by a number of translations of his treatise [62,63] and by its discussion in Heimann Breßler's Die Krankheiten des Kopfes und der Sinnesorgane (in three volumes, 1840) [64] and in Maximilian Joseph Chelius's (1794-1876) Handbuch der Augenheilkunde (1843) [65]. Especially Chelius' reputationhe had already published his highly influential Handbuch der Chirurgie (1822-1857; 8 editions and 11 translations) at that time and was one of the best-known surgeons in Europe (Fig. 4)  Sichel's idea also found its way into the French literature. While Auguste Théodore Vidal (1803-1856) in his 1840 Traité de pathologie externe et de médecine opératoire [67] limited himself mainly to reproducing Sichel's classification of amaurosis, which included spinal amaurosis as one category among many, later authors mocked Sichel's (and other German authors') indeed somewhat scholastic and, to some extent, speculative system of classifications of amauroses ("l'imagination puissante mais aventureuse de l'Allemagne émet toute une classification d'amauroses, dont l'interminable nomenclature semble nécessiter des prodiges d'analyse et ne nécessite que des prodiges de mémoire" [the powerful yet adventerous German imagination emits a whole classification of amauroses, the endless nomenclature of which seems to require prodigies of analysis yet only requires prodigies of memory] [68]) or tried to put the relevance of coexisting spinal diseases into perspective ("Il ne faut pas s'exagérer l'importance de l'intervention de la moelle épinière sur la production des phénomènes amaurotiques" [The importance of the contribution of the spinal cord on the generation of amaurotic phenomena should not be exaggerated] [22]). While Charles Deval in his Traité de l'amaurose ou de la goutte-sereine (1851) widely followed his teacher Sichel in his definition of spinal amaurosis ("[d] es phénomènes amaurotiques viennent-ils s'ajouter aux symptômes de la myélite, du ramollissement de la moelle épinière, d'une altération quelconque de celle-ci ou de ses annexes" [amaurotic phenomena come along with the symptoms of myelitis, softening of the spinal cord, any alteration of the spinal cord or its annexes] [22]), he also critically added for consideration that the validity of some aspects of Sichel's concept were "loin encore d'ètre démontrée" [far from being proven].
In England, the concept of spinal amaurosis was introduced by Edward Octavius Hocken, a young ophthalmologist, whose reports on patients with amaurosis and spinal cord disease we recently re-discovered [69]. Whether Hocken was familiar with Sichel's works or whether he developed the concept independently in parallel with Sichel we cannot know with certainty. However, in his Pathology and Treatment of Amaurosis, published in The Lancet in 13 parts between March 1841 and April 1842 [70][71][72][73] and, in parallel, in The London Medical Gazette [74], Hocken referred solely to his own experience, which, as he was convinced, "proved the occasional dependence of imperfection or loss of vision solely on spinal disease"a dependence that, as he pointed out, "seems to have been overlooked by all previous ophthalmological writers".
'Spinal amaurosis' in the second half of the 19th century The term 'spinal amaurosis' remained in use throughout the entire 19th century. However, it was not always used to refer to the association of acute amaurosis and myelitis. Already Sichel had adverted to debauchery (i.e. venereal diseases) and Chelius to tabes dorsalis as possible causes of spinal amaurosis. Von Walther, in his Die Lehre vom schwarzen Staar und seine Heilart. Pathologie und Therapie der Amaurose [75], even limited the symptoms of 'spinal amaurosis' to those of tabes dorsalis. This said, it should be taken into account that von Walther, a trained ophthalmologist, might well have intermingled cases of tabes dorsalis and acute myelitis as can be seen from his definition of the former disease: "Die Symptome der Tabes dorsalis (...) Paralysis oder Paresis der untern Extremitäten; Incontinenz des Urines." [the symptoms of tabes dorsalis (…) paralysis or paresis of the lower extremities; urinary incontinence].
In the second half of the century, more and more authorsincluding such eminent ophthalmologists such as Carl Ferdinand Ritter von Arlt (1812-1887) [76], his pupil von Graefe [77,78], and some of Graefe's own pupils such as Richard Liebreich   [79]), as well as parts of the medical press [80]would use the term preferentially or exclusively in that particular sense (referred to as 'amaurosis tabidorum' by some authors [64,[81][82][83][84]), while the association of amaurosis with myelitis suggested by Sichel, de Villards and Himly would be neglected. Eugène Follin (1823-1867) in his Leçons sur l'application de l'ophtalmoscopie (1859/ 1863), the earliest French work devoted entirely to the ophthalmoscope [85], Dixon (1855/1866), in his wellknown ophthalmological textbook [86,87], Richard Förster (1825-1902), in his chapter on neurological diseases in Gräfe und Saemisch's Handbuch der gesammten Augenheilkunde (1876), the most comprehensive ophthalmological textbook of the time [88], and Ernst Pflüger (1846-1903), in his oft-cited 1878 review on optic neuritis in Gräfes Archiv für Ophthalmologie [89], did not even mention the concept of 'spinal amaurosis'. Similarly, many of the most distinguished exponents of the newly emerging discipline of neurology were not aware of the association: Neither Moritz Heinrich Romberg (1795-1873) in his Lehrbuch der Nervenkrankheiten des Menschen (several editions between 1840 and 1857), which contains a widely read chapter on amaurosis, nor Karl Ewald Hasse (1810-1902; teacher of Robert Koch and Wilhelm Wundt) in his comprehensive chapter on myelitis in Krankheiten des Nervenapparates (1855), another reference textbook of the time, refer to the syndrome. Ernst von Leyden (1832-1910) had still pointed to the presence of amaurosis in spinal cord diseases other than tabes dorsalis in his 1863 monograph Die graue Degeneration der hinteren Rückenmarksstränge in a footnote: "Es ist übrigens nicht unwahrscheinlich, dass sich die Atrophie des Opticus auch zu anderen Rückenmarkskrankheiten [als dem Tabes dorsalis] gesellen kann" [It is, by the way, not improbable that optic nerve atrophy may be associated with spinal diseases other than tabes dorsalis] [90] but did not make any mention any more of that association in his renowned Klinik der Rückenmarkskrankheiten, which appeared in two volumes in 1874 and 1876 [91]. Missing out on Sichel's, Carron du Villards', and Pétrequin's observationsas well as on the early reports by Edward Hocken (1820-1845) [69], Giovanni Battista Pescetto (1806-1884) [31], Christopher Mercer Durrant (1814-1901) [92], Jacob August Lockhart Clarke (1817-1880) [93], and Thomas Clifford Allbutt (1836-1925) [94] recently rediscovered by us -, Wilhelm Heinrich Erb (1840-1921), considered by many the founding father of German neurology, would declare in 1878 in his reference book on diseases of the spinal cord, "von Störungen der Sehnerven ist bei der acuten Myelitis nichts bekannt" [nothing is known of disturbances of the optic nerves in patients with acute myelitis] [95]. In 1880, on the occasion of his report on the first known German patient with NMO (published in parallel by Erb and Philipp Stephan, a Frankfurt ophthalmologist [96]), Erb would still claim that such an association had been widely unknown before [97]. Similarly, William Gowers (1845Gowers ( -1915, then the most eminent British neurologist, would state, in an address on Eye symptoms in diseases of the spinal cord delivered before the Ophthalmological Society, in 1883 [98,99]: "I have ( …) never [seen optic nerve damage] in ( …) myelitis". Also in Charcot's lectures we were not able to find any reference to cases of spinal amaurosis other than tabetic amaurosis.
However, the association of amaurosis and spinal cord disorders other than tabes had never been completely forgotten. Christian Georg Theodor Ruete  Abercrombie (1780Abercrombie ( -1844 in 1828, which is the first known account in the Western literature of visual loss in a patient with spinal cord inflammation but no brain pathology [104]; in that patient, no signs of tabes or locomotor ataxia were noted. Similarly, Carl Stellwag von Carion (1823-1904) in the 1861/1862 edition (as well as in later editions) of his highly successful Lehrbuch der praktischen Augenheilkunde [105] (at least four English editions would appear between 1868 and 1873; also translated into Italian and Hungarian) reminded of "Amaurosis spinalis". He stressed that not only tabes dorsalis but various disorders of the spinal cordincluding spinal cord inflammationmay be associated with amaurosis: "Jedenfalls ist in Rechnung zu stellen, dass entzündliche Affection [des Sehnerven] (...) und dieses oder jenes Rückenmarksstranges fast gleichzeitig und ganz unabhängig von einander gar nicht selten beobachtet wird." [In any case, we must keep in mind that almost simultaneous and independent inflammatory affection [of the optic nerve] (…) and of this or that spinal nerve fibre bundle is not infrequent] [105], though he conceded that the exact relationship of the two affections remained a riddle: "Der Zusammenhang des Spinalleidens mit dem schwarzen Staare ist bisher noch ganz dunkel geblieben" [The relationship of spinal disease with amaurosis is still totally obscure] [105]. Eugène Bouchut (1818-1891), whose textbook on ophthalmoscopy (entitled Du diagnostic des maladies du système nerveux par l'ophthalmoloscopie 1866) appeared 5 years before Allbutt's [106] and 13 years before Gower's [107], was convinced that the introduction of the ophthalmoscope would make it possible to diagnose diseases not only of the brain but also of the myelon: "Si la découverte de cet instrument a été l'origine de progrès importants pour l'étude des maladies de l'oeil, sachons qu'il peut être la source de progrès non moins précieux dans le diagnostic des maladies cérébro-spinales en nous donnant le moyen de découvrir au travers de l'oeil les altérations qui se produisent dans les différentes parties du cerveau et de la moelle" [If the discovery of this instrument (i.e. the ophthalmoscope) was the origin of significant progress in the study of diseases of the eye, it can be the source of no less valuable progress in the diagnosis of cerebro-spinal diseases by giving us a means to discover through the eye alterations that are produced in different parts of the brain and spinal cord] [108]. He explicitly distinguished amaurosis associated with locomotor ataxia (i.e. tabes dorsalis) from amaurosis in patients with ordinary myelitis. In Édouard Meyer's (1838-1902) influential textbook Traite pratique des maladies des yeux (four French editions between 1873 to 1895), which is based on his 1863 lectures at the Ecole pratique of the University of Paris and which appeared in German, English, Russian, Polish, and Spanish translations throughout the 1870s and 1880s, we read: "L'amaurose spinale se rencontre surtout dans les cas de dégénérescence grise des cordons postérieurs (tabès dorsalis), plus rarement dans la myélite des cordons latéraux" [Spinal amaurosis occurs mainly in cases of grey degeneration of the posterior cords (tabes dorsalis), more rarely in the myelitis of the lateral cords] [109]. Meyer also explicitly mentioned paraplegia, a manifestation not typically associated with tabes dorsalis (which is primarily a slowly progressive degenerative disease of sensory nerves and ganglia in the dorsal columns and sensory roots [110]), among known accompaniments of amaurosis. Meyer, a pupil of von Graefe, carried on Sichel's work in Paris and purchased his clinic [111]. Theodor Leber, nowadays mostly remembered as the eponym of Leber's hereditary optic neuropathy, dedicated a whole chapter to the topic, entitled Spinalamaurose, in Graefe und Saemisch's Handbuch der Gesammten Augenheilkunde (1877). While he mainly dealt with the subject of tabes dorsalis in that chapter, he also pointed to the fact that other spinal disorders may rarely cause amaurosis ("doch ist Ihr Vorkommen auch bei anderen Rückenmarksleiden nicht selten"). However, of all the previously published cases, he was only aware of Allbutt's cases, incorrectly stating that "über Sehnervenleiden bei acuten Rückenmarksaffectionen ist [sonst] nur wenig bekannt "[about optic nerve diseases in spinal cord affection (otherwise) little is known] [112]).
The rare association of amaurosis with spinal cord disorders was also recognised by English-speaking authors: The 1854 edition (but not earlier editions) of William Mackenzie's (1791-1868) A practical treatise on the diseases of the eye [113], one of the leading British textbooks on ophthalmology of the period, listed "diseases of the spinal cord" among possible "complications of amaurosis", thus suggesting a causal, pathogenetic relationship between the two events. (When studying Mackenzie's work, the reader should carefully note that the term 'paralytic amaurosis' used elsewhere in that book refersfollowing Beer's nomenclatureto cases of amaurosis associated with paralysis of eye muscles but not to such of amaurosis and paralysis of the limbs.) The term 'spinal amaurosis' was also kept alive by von Graefe's former assistant John Soelberg Wells (1824-1879) in his Treatise on Diseases of the Eye (1869 [114]; 3rd edn. 1873; 4th American edn. 1883; translated into French and German), which according to Plarr's Lives of the Fellows of the Royal College of Surgeons of England embodied "the best teaching and practice of Continental and British practice" of its decade and "the standard textbook on the subject". Wells maintained the view that, "The affection of the optic nerve in diseases of the spine is probably due to a lesion of the great sympathetic through its communication with the anterior roots of the spinal nerves". Similarly, Boston-based Henry Angell (1829-1911) would refer to "spinal disorders" in general rather than to tabes in particular in his 1870 Treatise on diseases of the eye [115], which went through seven editions between 1870 and 1891, and John Phillips, one of the instigators of the use of the ophthalmoscope in the USA [116], explicitly referred to amaurosis "complicated with (…) paraplegia" and with inflammation of the spinal cord in his 1869 textbook Ophthalmic surgery and treatment [117]. Notably, Phillips also mentioned "morbid changes about the medulla oblongata" as a "precursor of spinal amaurosis". Today, we know that NMO is rather frequently associated with brainstem lesions and that lesions of the medulla oblongata, typically causing intractable vomiting and/or hiccups, indeed often herald the onset of NMO [1,10,118]. Finally, Allbutt would take up the term in his textbook On the use of the ophthalmoscope in diseases of the nervous system and of the kidneys (1871) [106], in which, tucked away in the appendix, a more detailed description can be found of the patient with ON and acute myelitis that he had briefly mentioned in his famous lecture on the ophthalmoscopic signs of spinal disease, published the year before in the Lancet [94,119]. Had Allbutt used the term 'spinal amaurosis' in the Lancet article, which most likely had a much broader readership than his book and which has been so often (falsely) cited as the first mention of a patient with NMO, the term might not have fallen into oblivion. Finally, Henry Swanzy (1843Swanzy ( -1913 in his Handbook of the diseases of their eye and their treatment (1892/1915) also refers to "spinal amaurosis" [120,121]. Swanzy also counted insular sclerosis (i.e. MS) among its possible causes.
By the end of the century, the concept of NMO was widely accepted. Von Leyden atoned for his former omission in his monumental, 970-page standard textbook Die Erkrankungen des Rückenmarkes und der Medulla oblongata (1897; together with Johannes Goldschneider ) [139], in which he acknowledged the co-occurrence of myelitis and ON. As a limitation, however, Leyden referenced Erb's 1879 case [97] as the first description ("zuerst von Erb [beobachtet]" [first observed by Erb]), overlookingjust like Erb before himall earlier cases, including those reported by Sichel, Carron du Villards, Pétrequin, Hocken [69], Pescetto [31], Durrant [92], Clarke [93] and Allbutt [94]. During the following twentieth century, NMO would often be mistaken as a variant of MS. It is therefore of note that von Leyden pointed to the fact that spinal amaurosis is distinct from MS in some cases according to autopsy findings: "In einigen Fällen von Myelitis, auch solchen, bei welchen durch die Autopsie die Diagnose gesichert worden ist, ist Neuritis optica beobachtet worden (...) Es scheint sich vorzugsweise, aber doch nicht ausschließlich, um die disseminierte Form der Myelitis zu handeln "[In some cases of myelitis, including cases in which the diagnosis was confirmed by autopsy, neuritis optica has been observed (…) It seems these cases represent preferentially but not exclusively the disseminated form of myelitis] [139]. Etiologically, von Leyden considered those cases to be of mostly postinfectious origin. In the second, revised edition in two volumes (1902/1904), he would repeatedly refer to the coincidence of acute myelitis and ON (citing a number of then recent clinical and pathological studies [125,129,134,140,141] on NMO), which he recognised as an important differential diagnosis of neurosyphilis.
Sichel's primacy in coining the term 'spinal amaurosis' Considering that there is rarely such thing as a new idea, we performed a careful analysis of much of the early 19th century French, German, English and Italian ophthalmological literature that preceded Sichel's report, including, among others, the works of Trnka von Kržowitz (1781/1791) [142,143], Beer (1791 [144], 1792 [145,146] and 1813/1817 [147,148] [160]) and Rognetta (1839 [172]), to verify Sichel's primacy in coining the term 'spinal amaurosis'. This extensive search revealed no earlier instance of the use of the term.

Amaurosis and spinal cord disease before Sichel
While the term 'spinal amaurosis' thus seems indeed to have been introduced by Sichel, the fact that optic nerve and spinal cord disease may occasionally co-exist had not completely escaped the attention of ophthalmologists and pathologists: Grundriss der allgemeinen Augenheilkunde (1834) [169], which antedates Sichel's Traité de l'ophthalmie by 3 years, had already referred to "Lähmung des ( …) Rückenmarks" [paralysis of the spinal cord] as a possible cause of amaurosis. Unfortunately, however, Andreae did not go into more detail regarding the nature and causes of that 'paralysis'. (The use of the term 'paralysis' in regard to nervous system structures such as the spinal cord or the optic nerve rather than to voluntary muscles was not uncommon but had already been subject to criticism by contemporaries of Andreae; see, for example, the 1825 edition of Travers's A synopsis of the diseases of the eye, and their treatment, the earliest systematic treatise in English on diseases of the eye: "To apply the term paralysis to a nervous tissue, is, to say the least of it, a misnomer. A muscle may be paralysed from pressure or injury of the nerve which supplies it, or the part of the brain or spinal marrow whence that nerve is derived; but there is certainly nothing in the nerve itself sufficiently analogous to muscular structure or function to apply the term 'paralysis' to both"  [207]) and should become a matter of heated debate in the 1870s following the introduction of the ophthalmoscope and, in particular, the publication of Wharton-Jones' Failure of sight from railway and other injuries of the spine and head in 1869, and would be harshly criticised by Gowers [206]. The discussion about the exact mechanism by which spinal cord injury could possibly cause optic nerve damage (nutritional damage to the optic nerve resulting from damage to the spinal origins of the sympathetic nerves vs. ascending meningeal 'irritation' or meningitis) would later on also determine the discussions on pathomechanisms underlying the simultaneous or successive occurrence of myelitis and ON in NMO for a long time. g. In addition, we came across a number of case reports compatible with (and in at least one case highly suggestive of) NMO, all of which predate Sichel's 1837 report by years or decades. While these reports (to be presented in the following sections), some of them authored by eminent figures of early 19th century medicine in Great Britain, do not challenge Sichel's primacy in coining the term 'spinal amaurosis', they further demonstrate that the syndrome itself had already been recognised by ophthalmologists and pathologists before Sichel.
First, we would like to draw the reader's attention to a very early case series of patients with 'gutta serena' (a historic term for blindness) and para-or tetraplegia collected by Matthew Baillie (1761-1823) (Fig. 5), a Scottish-born British physician, anatomist and pathologist, shortly before his death in 1823. Baillie's notes were made known to the medical public only posthumously in 1826. Baillie, who had been made Fellow of the Royal Society (FRS) in 1790, was nephew and pupil to the renowned anatomists John Hunter, FRS (1728-1793), and William Hunter, FRS (1718-1783), was educated at Oxford, and is nowadays mostly known for his celebrated Morbid anatomy of some of the most important parts of the human body (1793) [208,209], the first systematic textbook on pathology in English and the first to make pathology a subject in itself (see ref. [210] for a biographic sketch). He is also considered to have given the first description of a patient with situs inversus viscerum. Finally, readers interested in the history of medicine may be familiar with Baillie as the last owner of the famous gold-headed cane originally possessed by no less figure than Oxford's great benefactor John Radcliffe (1652-1714) [211].
In 1820, in a paper entitled Some observations on paraplegia in adults and published in the sixth and last volume of the Medical Transactions of the London College of Physicians [212], Baillie proposed that cerebral lesions could cause paraplegia ("[a condition] in which the lower half of the body is more or less impaired in its nervous power"). As an example for brain involvement in patients with paraplegia, he referred to patients with concomitant optic nerve damage (anatomically, the optic nerve is part of the brain): "Sometimes the sight of one eye is almost entirely lost, and its pupil appears dilated as in gutta serena" [212].
While the hypothesis of cerebral disease being capable of causing paraparesis was considered acceptable by many of Baillie's contemporaries, his opinion that "if the spine has not suffered outward violence, paraplegia most commonly (our emphasis) depends on a disease of the brain itself" was met with reservation by some and openly rejected by others (see, for example, reference [213]: "[I]t is not a little remarkable that one whose opportunities and whose powers of observation were so good, was led to adopt such views. Can it be accounted for by supposing that, in those days, the spinal cavity was even less opened after death than at the present time?"), all the more as Baillie gave results from only one dissectiona case of paraplegia associated with extensive effusion in the ventricles and membranesin support of his theory. He later (in a brief paper dated June 1822 and posthumously published in an edition of his works arranged by James Wardrope in 1825 [214]) added three more cases, one caused by "tumours in the brain", one in which "many of the arteries of the brain were found ossified", and another in which "a large quantity of water" was found in the ventricles of the brain (but also "some in the theca vertebralis").
However, Baillie obviously felt that these cases might not suffice to justify his far-reaching claim and continued to collect further cases, which he briefly described in a short letter he marked Short memoranda of cases of paraplegia on the back and entrusted to a colleague, a Dr. Gregory (probably George Gregory [1790-1853], the grandson of William Cullen's [1710-1790] successor at Edinburgh, James Gregory [1753-1821], who had studied anatomy at the Windmill Street School under the tutelage of Baillie; Baillie and Gregory's father were friends from their early lives at Balliol College, Oxford [215,216]), to be produced should his opinions ever be particularly controversial [217]. All of those patients had developed amaurosis, which Baillie curiously took as sufficient proof that paraplegia was caused by cerebral disease as well. In 1826, these memoranda were read on the occasion of a meeting of the Westminster Medical Society (later printed verbatimthough now entitled Facts relative to the paraplegia (From a posthumous MS. of the late Dr. Baillie) [217]). In these memoranda, which sparked a lively discussionincluding fair and objective contributions [217][218][219][220][221] as well as vituperative personal attacks [222] -Baillie briefly mentioned a series of five cases of amaurosis or impaired vision and para-or tetraparesis: 1. "Another gentleman, in paraplegia, has gutta serena of his left eye; had great weakness in his arms, with indistinctive feeling, so that he said he could not distinguish, by the touch, shillings and sixpences from each other." 2. "A lady, in paraplegia, had impaired vision, severe headache, and weakness in her arms and hands." 3. "A clergymen had gutta serena of one eye along with paraplegia." 4. "A gentleman had a temporary gutta serena, and an occasional dropping of one eyelid, with paraplegia." 5. "A nobleman had the vision of both eyes very much impaired in paraplegia from (sic!) gutta serena, but his affection at length a good deal subsided." While it can retrospectively not be proven that these patients had indeed AQP4-IgG-or MOG-IgG-related NMO, Baillie's report again demonstrate that the peculiar coexistence of optic and spinal symptoms that characterises NMO, i.e. the isolated affection of two sites not apparently connected in terms of anatomy or function, attracted the attention of physicians much earlier than previously thought. With the exception of headache in one and occasional dropping of an eyelid in another patient, there was no indication for brain disease other than possible optic nerve involvement, whichthough the absence of brain symptoms does not per se preclude early MSrenders a diagnosis of MS less likely.
As a limitation, it is unclear whether symptoms developed acutelyas typically seen in NMOin Baillie's patients. In consequence, tabes dorsalis cannot be completely ruled out. Some of the patients included by Devic and Gault in their seminal study of early NMO cases [44] did indeed have a history of syphilis, a disease that was highly prevalent throughout the 18th and 19th centuries. However, no such history was related by Baillie, nor did he report typical clinical symptoms of tabes dorsalis such as lancinating pain or paralytic dementia. Moreover, tabes dorsalis primarily affects the posterior columns; accordingly, paraplegia is not a typical manifestation. Finally, symptoms were transient in case 4 und substantial recovery was noted in case 5, in contrast to the chronic progression typically seen in neurosyphilis.
Regarding treatment, Baillie recommended, among other 'remedies' of the time, electrotherapy: "In one case I found that a good deal of benefit was derived from electric sparks being drawn from the lower limbs". Electric treatment of patients with NMO would still be tried in the late 19th century [123]. To the best of our knowledge, Baillie's report has never been cited again.
Bilateral amaurosis following a violent attack of 'pleuritis dorsalis': James Ware (1795) To assess whether Baillie and Portal were really the original 'giants' on the shoulders of which later generations stood (or sat, as the windows of the Southern transept of Chartres cathedral suggest [223]), we undertook to expand our search to the most important medical textbooks of the second half of the 18th century. However, any study of the history of NMO is complicated by the fact that modern ophthalmological and neurological terminology deviates from that used in the 18th and early 19th century.
Blindness was not always referred to by the name of 'amaurosis' at that time but also by a plethora of other terms, including 'gutta serena', 'goutte-sereine', 'suffusio nigra', 'schwarzer Star' or 'amblyopia' (weakness of sight).
Of the latter termsmost of which are little known todayparticular importance attaches to 'notaeomyelitis' and 'pleuritis dorsalis'. Under the title of 'notaeomyelitis' (or, more precisely, its Italian equivalent 'noteomielite') Pescetto's reportone of the earliest detailed reports on a patient with NMOwas published [225]. As pointed out by us in Part 1 of this article series [5], the term 'notaeomyelitis' was originally introduced by the Austrian physician Johann Valentin Hildenbrand (1763-1818), the eponym for Hildenbrand's disease (i.e. typhus), in his textbook Institutiones practico-medicae (Vol. III [published posthumously in 1822], p.97ff) to distinguish the inflammation of the medulla spinalis or d o r s a l i s (termed μυελος ν ω τ ι α ι ο ς by Hippocrates) from that of other types of marrow such as the bone marrow (osteo-myelitis) or the cerebral white matter (encephalo-myelitis). The use of that rather unusual term may well have contributed to the fact that Pescetto's early account of NMO went unrecognized for the next 150 years. Of special note, Hildenbrand was also the first to use the term neuromyelitis (or more precisely, nevromyelitis). However, by that term Hildenbrand was referring not to the co-occurrence of ON and myelitis, but to inflammation of the pulpa nervorum as opposed to that of the vagina nervorum (nevrilemmatitis).
We believe it was in that sense that the term was applied also by James Ware (1756-1815) (Fig. 6) in his Enquiry into the causes which have most commonly prevented success in the operation of extracting the cataract; with an account of the means by which they may either be avoided or rectified; to which are added observations on the dissipation of the cataract, and on the cure of the gutta serena [228]), which first appeared in print in 1795, i.e. 18 years before Harleß would propose substituting 'pleuritis dorsalis' by the new term 'myelitis'. In that treatise, Ware describes the case of a 45-year-old woman who experienced an attack of a "violent pleuritic disorder" which reduced her strength so much that she "became even unable to turn herself in bed without assistance". One month after onset, she suffered from "violent pain in the left eye" and headache, which was soon followed by considerable loss in vision in that eye, which within 5 days became completely amaurotic. Three weeks later, she developed similar pain in the right eye, again followed by a decline in vision leading to almost complete blindness after 2 to 3 weeks and complete blindness after 1 month. At that time the violent eye pain ceased. Upon examination there were no signs of inflammation of the external eye structures, but the pupils of both eyes were found to be dilated with no apparent light reaction. The patient was removed into the country site, where she recovered her strength but not her sight. Treatments included vitriolic aether (i.e., diethyl ether), applied several times per day to both eyes, Peruvian bark, and other 'remedies'. Of particular note, however, she was in addition treated by electrotherapy: "a strong stream of the electric fluid was to have been applied to the eyes" and continued for ten to fifteen minutes. While Ware was confident that this treatment had been "not a little serviceable on every trial", it had to be ceased due to a little accident: "the person employed in the business of the electricity, being unacquainted with the mode of applying the stream", accidentally substituted for it "electric shocks (…) through the head". In consequence, the patient was "electrified" only three times, and "in this way". Within 3 months time she recovered enough vision to enable her to read common letters with her left eye and to see all larger objects with the right. Ware included that report also in the second and third edition of his treatise, which were published in 1804 and 1812, respectively, under the abridged title Observations on the cataract and gutta serena.
The subacute onset, the close temporal relationship of the two events, the intensity of the attacks (leading to almost complete paresis and complete visual loss within short time), the symptoms accompanying visual loss (dilated pupils rather than tabetic miosis, violent eye pain), the absence of signs of cerebral disease, and, finally, the partial recovery, all strongly argue against tabes (though other, rare complications of syphilis, a history of which was not mentioned by Ware, cannot be formally excluded) but strongly support a diagnosis of NMO in this patient.
Interestingly, the patient, a women of 45 years (the median age of onset of AQP4-IgG-positive NMO, which mainly affects women, is 40 years [1]), had been for many years "subject to frequent returns of rheumatic affections in different parts of the body". Given that AQP4-IgG-positive NMO has been demonstrated to be frequently associated with connective tissue disorders such as systemic lupus erythematosus, Sjögren's syndrome, or rheumatoid arthritis in a substantial number of cases [1,59,229], the latter fact does not argue against but rather supports a diagnosis of NMO. As a limitation, the exact nature of this patient's rheumatic affliction remains elusive, since Ware does not provide us with more detail. Ware, a pupil of Jonathan Wathen (1728-1808) and contemporary of Beer, is considered one of the founding fathers of modern ophthalmology in Great Britain [230][231][232]. He was admitted a Fellow of the Royal Society in 1802the first "bare oculist" in the history of the society. The Dictionary of National Biography underlines Ware's contributions to establishing ophthalmology as a scientific subject in Britain: "It is the peculiar merit of Wathen and of his pupil Ware that they elevated ophthalmic surgery from the degraded condition into which it had fallen. Originally a branch of general surgery, but always invaded by quacks, it fell into dishonest hands, from which the disinterested efforts of men like Ware first rescued it" [230].
However, electrotherapyfirst applied by Christian Gottlieb Kratzenstein (1723-1795) and Johann Gottlob Krüger (1715-1759) in Germany in the 1740s and much promoted by John Wesley (1703-1791), the founder of Methodism, in Great Britainwhich both Baillie and Ware commended to their patients, was itself prone to charlatanry. The period following Luigi Galvani's (1737-1798) discovery in 1780 and the debate with Alessandro Volta (1745-1827), and especially the years around the turn of the century directly preceding the appearance of Ware's report, had seen the publication of numerous books on the topic (e.g., Ritter 1800; Reinhold 1797/98; Grapengießer 1801, Augustin 1801, Struve 1802, Hellwag 1802, Sternberg 1803, Clarus 1802/3, Weber 1802, Eschke 1803, Reinhold 1803), and many of the therapeutic recommendations made at the time were insufficiently (criticised as "grobempirisch" as early as 1816 [233]) or not at all evidence-based. Figure 7 shows an apparatus used to treat amaurosis (taken from Christian Heinrich Ernst Bischoff's [1781-1861] Commentatio de usu galvanismi in arte medica speciatim vero in morbis nervorum paralyticis in which he propagated the use of galvanism for treating neurological diseases, including paresis and amaurosis as well as rheumatic disorders [234]).

Joseph Warner (1773): severe amaurosis and paraparesis
In his Description of the human eye, and its adjacent parts; together with their principal diseases, and the methods proposed for relieving them (1773) [235], which Beer highly esteemed despite its brevity, and which Hirschberg consideredas he felt it was based on practical knowledgethe first useful ophthalmological textbook written in the English language, the Antigua-born Joseph Warner (1717-1801) (Fig. 8), then first physician to Guy's Hospital (a position he held for 40 years) and Fellow of the Royal Society [7,236], reported the following case of severe bilateral amaurosis with concurrent paraplegia but apparently no signs or symptoms of brain involvement: "In the instance of a young gentleman, whom I several times visited on a similar occasion, his defect of sight was accompanied by a palsy of the lower limbs (…) the fibres of the iris had no visible motion; the humours of the eyes preserved their natural transparency; his sight was not totally lost, but it was become so much impaired, as to be of little or no use to him; and it seemed to be affected in a degree pretty nearly equal to that of his legs and thighs." This could well be the earliest description of a case of NMO known so far. The words "several times on a similar occasion" even suggest a relapsing course of disease. However, it is an obvious limitation that Warner specified neither whether the symptoms were of acute onset nor whether any other signs or symptoms of myelitis were present. Accordingly, considerable uncertainty regarding the exact diagnosis must remain.
Given its high prevalence, syphilis would be a potential differential diagnosis. It is therefore of particular interest that Warner, who related this case as an example for his more general observation of amaurosis not always (although mostly) being followed by mydriasis, gives us the following additional information: "in [this patient's] eyes the pupils were contracted to at least half their natural size". This could well be a very early description of syphilitic miosis, which would later be analysed in detail by Douglas Argyll Robertson (1837Robertson ( -1905, von Graefe's famous Scotish pupil (cf. Argyll Robertson's seminal 1869 articles on the topic [237,238], in which, however, he acknowledges having learned about the phenomenon already at Robert Remak's [1815-1865] clinic in Berlin and in which he also referrs to earlier descriptions by Romberg [to be found in vol. 1, dev. 2, of his Lehrbuch der Nervenkrankheiten des Menschen, first edn. 1840-1846 [239], English translation 1853 [240]], by Armand Trousseau  [see the revised chapter on locomotor ataxia included in the third volume of his Clinique médicale de L'Hotel-Dieu dè Paris, first edn. 1865 [241]; English translation 1867 [242]], and by Stellwag von Carion in his Lehrbuch der Augenheilkunde [first edition 1861/1862 [105]; English translation 1868]; see also [243,244]).
On the other hand, disturbances in pupillary function, including miosis, have been repeatedly reported also in AQP4-IgG-positive NMO [245][246][247][248][249], and the periaqueductal area, believed by some authors to contain the mostly likely site of damage in patients with Argyll Robertson syndrome, is also a site of high AQP4 expression and, accordingly, not infrequently a site of inflammation in AQP4-IgG-positive NMO [9,250]. Finally, brainstem involvement is also a common feature in MOG-IgG-positive NMO [10], and periaqueductal lesions have also been reported in this condition. The presence or absence of miosis thus might not be a sufficient criterion for distinguishing neurosyphilis and autoimmune NMO.
The relationship of neurosyphilis and NMO is complex. NMO was originally defined as a syndrome rather than a pathogenetically defined disease entity [5,44]. In consequence, the presence of syphilis would per se not necessarily have precluded a diagnosis of NMO. As mentioned above, some of the patients included by Devic and Gault as cases of NMO in their disease-defining review [44] in fact had a history of syphilis. Whether this was a simple coincidence owing to the high prevalence of syphilis at that time (only a minority of patients affected by primary lues will later develop neurosyphilis) or whether NMO was caused by syphilitic lesions in these cases is unknown. Apart from tabetic degeneration of the optic nerves and posterior columns, neurosyphilis can cause spinal pachymeningitis, meningomyelitis, and spinal endarteritis, which have at least the potential of causing paraplegia (see also Erb's 1902 London lecture On spastic and syphilitic paralysis [251]). Similarly, ON is counted among the possible (though very rare) complications of neurosyphilis. The incidence of neurosyphilis, one of the most common neurological diseases of the 18th and 19th century, has significantly declined since the introduction of penicillin as a treatment for lues. Neurosyphilis is now a very rare diagnosis. Accordingly, in recent decades only a few patients with NMO and syphilis have been reported in the Western medical literature, one of whom was tested for NMO-IgG but was negative for that marker [252,253].
Alexander (1827), Marcé (1845/1847) and Bradley (1818) Mainly to avoid remonstration by the informed reader, who may be aware of these cases, but also for the sake of completeness, we would like to briefly mention three further, though more doubtful, reports of possible NMO that appeared in The Lancet in 1827; in a small local journal, the Journal de la Section de médecine de la Société académique du département de la Loire-inférieure, in 1847; and in a medical textbook entitled On some varieties of spinal disease, published in 1818.
The first report, published by John Alexander, a Manchester-based physician, describes the case of a 12year-old girl [254]. While the title of the article ("On intermitting paraplegia, combined with amaurosis") is highly suggestive of NMO, the case itself is very unusual in that the onsets of paraplegia and amaurosis were preceded or accompanied, respectively, by chorea-like disease, fits that "assumed the appearance of epilepsy" and "occasional phrensy inducing her to run out of the house, and proceed until exhausted". Moreover, rubbing a little ointment in the neck, as a placebo, was followed by rapid recovery, rendering the diagnosis of a functional disorder at least conceivable. The term "spinal amaurosis" was not used. In the medical literature of the 19th century, "hysteric" amaurosis, paraplegia or tetraplegia is a recurrent topic. On the other hand, it should be noted that epileptic seizures as well as psychiatric symptoms have been reported both in AQP4-IgG-positive and in MOG-IgG-positive NMO, in particular in children [12][13][14][15].
The second report is part of a series of observations on "myélites spontanées qui se sont sporadiquement manifestées à Nantes à dater des derniers mois de 1845" by Germain-Auguste Marcé (1805-1859; see reference [255] for a biographical sketch) and describes the case of a male patient, an abbot, who, after exposure to cold and humidity, developed fever, spastic and painful sensorimotor paraparesis and, in addition, plegia of the left arm and right hand. The patient did not response to various treatment attempts, including bloodletting, vesicatories, strychnine and quinine. New sensory symptoms occurred about 2.5 months after onset, and at around 5 months after onset he experienced an acute attack which led to blindness in the left eye and complete tetraparesis, associated with slight dysarthria. The patient died a few days later. No autopsy was performed. During the entire duration of the disease there was no clinical evidence for supratentorial brain involvement, and "les idées étaient parfaitement lucides". A diagnosis of inflammation of the spinal cord with concomitant visual loss was made. While the onset of both AQP4-IgG-and MOG-IgGpositive autoimmune NMO is not rarely preceded by acute infection, [1,51] the fact that fever and sweating persisted for a considerable time throughout the course of disease is rather atypical and suggests a possible infectious aetiology in this case.
The third case is that of a 17-year-old girl who presented with visual deficiency, a mydriatic pupil with diminished light reaction, some weakness of the lower extremities, considerable pain when pressure was exerted on the third and fourth vertebrae, and vertigo [256]. However, the case is otherwise rather obscure and it is doubtful whether the patient had NMO. The author's final recommendation, made as early as 1818, thus seems to be of more importance to us than the case itself: "[it is necessary] both for the oculist and [the] general practicioner clearly to ascertain, in cases of dysopsia, whether the disease be not accompanied with constitutional symptoms dependent on some affection of the spine." Peltesohn (1886): early epidemiological data on the coincidence of myelitis and optic neuritis We would like to draw the reader's attention to a work which seems particularly noteworthy by virtue of being the first 'epidemiological' work on NMO. In an article seriesan excerpt from his dissertationon the causes and course of optic nerve atrophy, published in three parts in Zentralblatt für praktische Augenheilkunde in 1886 [126][127][128], the German ophthalmologist Nathan Peltesohn (1862Peltesohn ( -1942 related the case of a 36-year-old merchant with cervical myelitis, resulting in spastic tetraparesis, hyperreflexia, paraesthesia, girdle band sensation, and urinary and erectile dysfunction. The patient developed complete bilateral amaurosis; ophthalmoscopy revealed optic nerve atrophy. However, similar to some patients reported by Devic and Gault [44], the anamnestic information suggested a history of lues. An infectious aetiology can thus not be excluded. Interestingly, Peltesohn's patient was clinically examined by no lesser figures than Julius Hirschberg, still known to many as the author of the monumental Geschichte der Augenheilkunde, published in nine volumes between 1899and 1918(English edition 1982-1994, and the psychiatrist and neurologist Emanuel Mendel (1839-1907), teacher to Max Bielschowsky (1869Bielschowsky ( -1940 and Sigmund Freud . In this work, Peltesohn included for the first time statistical data (from Hirschberg's department) on the prevalence of myelitis in patients with optic nerve atrophy. Among 248 patients with "neuritic atrophy" of the optic nerve(s), i.e. atrophy following ON, only two also (0.8%) had myelitis; and among 248 patients with non-neuritic, "simple atrophy" of the optic nerve, two (0.8%) had "chronic myelitis" (all patients were seen between 1880 and 1885). Dr. Peltesohn, who continued to practice privately in Hamburg after withdrawal of the accreditation of Jewish doctors with the statutory health insurance companies in 1933, was murdered by the Nazis, together with his two sisters, in 1942, aged 78.

François Boissier de Sauvages de Lacroix: 'Amaurose rachialgique' (1763)
In the final two sections we return briefly to nomenclature (an important topic, since it helps to structure medical knowledge: '[w]hat is a disease before it gets a name?' [T. Jock Murray, Multiple Sclerosis: The History of a Disease, New York, 2005]) by referring to a curious contribution by François Boissier de Sauvages de Lacroix's (1706-1776) and, finally, by discussing the reasons for the change in terminology that was to occur at the turn to the twentieth century.
In 1763, de Sauvage (botanical author citation: SAUV), inspired by Linné's ground-breaking works -Systema Naturae (which, despite all the known limitations resulting from its artificial nature, marks the beginning of modern zoological nomenclature) had just appeared in 1758and Sydenham's attempts to use a 'binary' or 'binomial' system also for the classification of diseases, published his famous and monumental Nosologia methodica in ten volumes. Further nosological compendia were published by William Cullen (1710-1790), Whytt's predecessor in Edinburgh, between 1769 and 1785, and by Jean Baptiste Théodore Baumès (1756-1828) in 1806.
Interested in whether any of these authors was aware of amaurosis spinalis (a binomial term in itself, consisting of species and epithet), we came indeed across a very similar term in de Sauvage's nosology, namely that of 'amaurose rachialgique'. Considering that Hippocrates had referred to myelitis as 'lumbago' and Brera as 'rachialigitis lumbalis', 'amaurose rachialgique' clearly evokes the concept of 'spinal amaurosis'.
The term would later be adopted by Jean Baptiste Théodore Baumès (1756-1828), who is now mainly remembered for his iatro-chemical ideas, i.e. for pointing to the connection between medicine and chemistry (Essai d'un système chimique de la science de l'homme 1798). In both his Fondemens de la science méthodique des maladies (1801-1802) [257] and his Traité élémentaire de nosologie (1806) [258] we find the following entry: "Amaurose rachialgique; SAUV., cl. VI, ord. I, gen. IV (should be gen. V according to our edition of de Sauvages), esp. 14". However, when looking up the lemma 'rachialgia', there is no explicit reference to spinal inflammation, but exclusively to pain related to the spine: "Syn. From 'spinal amaurosis' to 'neuromyelitis optica' Given the clear and distinct character of the term 'spinal amaurosis', which follows an Aristotelian rule of definition, which requires that "definitio fiat per genus proximum et differentiam specificam", it isat least at first glancenot easy to understand why it would be replaced by 'neuromyelitis optica', a highly artificial term that purists might regard as linguistic barbarism [260], later on. All the more so because the term 'neuromyelitis'though this may not have been known to Devic and Gaulthad been in use before, with different meanings. In total, we came across at least six different usages of the term, listed here in chronological order: i. 'Nevromyelitis' denoting inflammation of the pulpa nervorum as opposed to inflammation of the vagina nervorum ('nevrilemmatitis'); in this sense the term was probably first used by Johann Valentin Hildenbrand (1763-1818) in his famous Institutiones practico-medicae (published posthumously by his son in 1822) and later adopted [261][262][263] but also criticised [264][265][266][267][268][269][270]; ii. 'neuromyelitis' as a synonym of 'inflammation of the spinal cord', as defined in the 1836 edition of the prestigious Dictionnaire de l'Académie Française and later adopted by others (see, for example, reference [271]:  [44] or, most recently, 'neuromyelitis optica spectrum disorders' [250].
We believe that several factors determined the rapid change in nomenclature that took part at the turn of the century: 1. As mentioned above, a shift in use of the term 'spinal amaurosis' towards being employed as a synonym of 'tabes dorsalis with amaurosis' took place in the second half of the 19th century. This parallels the growing interest in tabes following accumulating evidence for its being caused by syphilis, especially after the appearance of Fournier's De l'ataxie locomotrice d'origine syphilitique in 1875/1876 [275,276] and, a fortiori, after support for Fournier's hypothesis from Erb [277] (in disagreement with Leyden), Vulpian (dissenting in that regard with his former co-author on tabes, Charcot) [278] and Gowers [279]. This created a need for a new term that would better distinguish the two conditions. 2. This shift may be related to the rise in prevalence of neurosyphilis over the course of the 19th century suspected by some authors [280] or, at least, be a result of the much higher prevalence of tabetic amaurosis than of NMO at that time. While NMO remains a rare disease to the present day, syphilis was highly prevalent far into the 20th century. The Royal Commission on Venereal Diseases in 1916 concluded that the disease's prevalence "cannot [be] below 10 per cent of the whole population in the large cities" [281]. The high number of tabetic cases may thus have simply overshadowed or eclipsed the previous knowledge about rare cases of non-tabetic spinal cord and optic nerve damage reported throughout the 19th century. This may be true especially for neurology, a clinical discipline that was just emerging (Charcot's chair at the Salpêtrière, the first one worldwide dedicated specifically to diseases of the nervous system, was established only in 1882) and in the process of setting up its own discrete nomenclature. Many physicians dedicating themselves to diseases of the nervous system at the end of the nineteenth century may simply not have been aware of the previous ophthalmological literature on spinal amaurosis, especially that from the first half of the century, as is evident, for example, from the above-cited statements by Erb, Allbutt and Gowers. 3. The fact that the term 'spinal amaurosis' places the emphasis on amaurosis, which may not appropriately reflect both the great disease burden inflicted by spinal disease as well as the greater interest neurologists may attach to myelitis than to amaurosis in general and in particular may have attached to that syndrome at a time when great progress was being made in elucidating the physiology and pathology of the spinal cord. 4. Finally, the wording 'spinal amaurosis' can be understood as implying a causal relationship. However, the notion of spinal cord disease causing amaurosis (e.g. by procuring damage to the spinal sympathetic nerves) in patients with 'spinal amaurosis' was not proven and, accordingly, not accepted by all, including such influential figures as Erb, Allbutt and Gowers.

Discussion
Until very recently, most review articles and book chapters on NMO let the history of that syndrome begin with Devic and Gault'sindeed seminalreview De la neuromyélite optique aiguë [44], which appeared in 1894, and a brief reference to patients with acute myelitis and optic nerve damage by Thomas Clifford Allbutt in his 1870 article on the ophthalmoscopic signs of spinal cord disease was considered by many the first mention of a patient with possible NMO in the medical literature [94]. However, as we were able to demonstrate in a recent series of articles, Devic and Gault as well as Allbutt overlooked numerous earlier case of NMO, probably owing to the limited bibliographic means of the time. These reportsone published in The Lancet in 1865 [93] by the famous British neuroanatomist and neurophysiologist Jacob August Lockhart Clarke, eponym for the posterior thoracic nucleus or Clarke's column, one by the British physician Christopher Mercer Durrant in 1850 in the Provincial Medical and Surgical Journal [92], the precursor of the BMJ, one by the Genoese physician Giovanni Battista Pescetto in a local Italian journal in 1844 [31], and one by Edward Hocken, a young but already eminent British ophthalmologist, in The Lancet in 1841all clearly precede Allbutt's article, demonstrating that the concurrence of amaurosis and spinal cord damage so characteristic for NMO had caught the attention of physicians much earlier than previously thought.
The cases re-presented here, which date back to the late 18th and early 19th centuryand thus mostly antedate even the cases described by Clarke, Durrant, Pescetto and Hockenprovide additional evidence for the latter notion. In addition, we re-present the concept of spinal amaurosis, which, although the immediate precursor of NMO, was virtually forgotten. Of note, this concept was introduced not by neurologists but by ophthalmologists, who were, as shown here, the first to take note of the syndrome. The term 'spinal amaurosis' (and its variants) can be traced back, as demonstrated here, to the first half of the 19th century and remained present in the medical literature until the end of that century, from which timein parallel with the establishment of neurology as a new discipline with its own, distinct nomenclatureit was gradually superseded by 'NMO' and its variants ('neuro-myélite optique aiguë', first proposed by Devic and Gault 1894 [44, 133, 137]; 'acute optic neuromyelitis', anonymous 1903 [282,283]; 'neuromyelitis optica', Stransky 1904 [284]) and by corresponding eponymous designations ('maladie de Devic', first proposed by Peppo Acchiote 1907 [285][286][287]; 'Devic['s] disease'; 'Devic['s] syndrome'; 'Devic-Erkrankung'; 'Morbus Devic').
While the simultaneous occurrence of optic nerve and spinal cord damage was, as shown here, already noted in the late 18th and 19th century, the exact nature of the pathogenic relationship between the two affections remained elusive. While some authors believed that damage to the spinal sympathetic nerve might impair the blood supply of the optic nerve, others believed that amaurosis resulted from ascending meningitis, originating in the spinal cord and propagated to the optic nerves per continuitatem, or wondered about the possible existence of an unrecognized anatomical connection between the spinal cord and optic nerves. Only a few authors, among them Gowers and Erb, but also Gault and Devic, would (explicitly or implicitly) dismiss the idea of a direct link and maintain the view that the two sites were independently affected by a common causal factor. It would take more than two centuries for more light to be shed on the pathogenesis of NMO, when a role for antibodies was suggested based on histopathological findings by Lucchinetti and colleagues in 1999 [57] and when finally aquaporin-4, the most abundant water channel in the CNS, was revealed as target antigen in the majority of patients with NMO by colleagues in 2004/2005 [2, 47]. Recent studies suggesting differences in AQP4 expression levels and isoform composition between CNS regions [288] as well as the hypothesis of a higher permeability or vulnerability of the blood-brain barrier in some areas of the optic nerve and spinal cord [289][290][291][292][293] provide a potential rationale for the predominant involvement of the optic nerves and spinal cord in this disease, which had been a riddle for more than 200 years. Moreover, pathogenic autoantibodies to MOG have been identified in a subset of patients with AQP4-IgG-negative NMO [10,[50][51][52][53]. In some AQP4-IgGand MOG-IgG-negative NMO patients a beneficial effect of plasma exchange has been reported, suggesting a humoral pathogenesis and the presence of (so-far unknown) autoantibodies. Finally, rare cases caused by paraneoplastic autoimmunity (e.g. to CV2/CRMP5 [294]) or related to neurosarcoidosis have been identified. The pathogenetic heterogeneity in NMO suggested by these findings made changes in nomenclature and nosology inevitable: NMO is no longer considered a single disease butwell in line with Devic and Gault's original definition rather viewed as a clinical or clinicoradiological syndrome or phenotype shared by two or more pathogenetically distinct diseases. AQP4-IgG-positive NMO and MOG-IgG-positive NMO are now recognised by most authors as entities in their own right distinct from classic MS. Accordingly, diagnostic criteria different from those used to diagnose MS [295] have been proposed [54,250].
The origins of the concept of spinal amaurosis fall into the 'ophthalmological renaissance' in early 19th century France that followed the temporary yet substantial decline of ophthalmology in the aftermath of the French revolution and which was driven mainly by the immigration of ophthalmologists from abroad, such as Sichel, and the recognition of the important progress ophthalmology had meanwhile made outside France [295]. However, as shown here, very early cases of possible NMO can also be found in the works of physicians based in other European countries, including the United Kingdom, and the term 'spinal amaurosis' was soon adopted throughout the continent.
Intriguingly, spinal amaurosis or NMO, remained a topic of interest at some centres from the 19th century, with some interruptions, right up to the present time: Pétrequin, born in Villeurbanne, for many years held the prestigious position of "chirurgien-major" at the Hôtel-Dieu de Lyon, the very city that later became linked forever with NMO by the seminal works of Devic and Gault [44,133,137], which would appear 18 years after Pétrequin's death. Since then, the Lyon school of neurology has maintained a strong interest in NMO (e.g., [296][297][298][299][300][301]), partly driven by Christian Confavreux (1949-2013), a world-renowned specialist in MS, who married into the Devic family [302]. Chelius, on the other hand, who contributed greatly to popularizing the new concept of 'spinal amaurosis', was a graduate of the University of Heidelberg, where he was appointed professor of general surgery and ophthalmology in 1817. Chelius, rated by August Hirsch in his Biographisches Lexikon der hervorragenden Aerzte aller Zeiten und Völker as one of the most renowned and esteemed physicians and surgeons of his time in Europe [303], is not only considered the founding father of the Heidelberg school of ophthalmology but was instrumental in laying the foundations for the later high reputation of the Heidelberg medical school [304]. Around 25 years after Chelius' book appeared in print, Wilhelm Erb became a professor at Heidelberg, where he laterafter an intermezzo in Leipzigwould succeed Nikolaus Friedreich (1825-1882), eponym of Friedreich's ataxia, as head of department. It was in 1879 that Erb published the first German report on a patient with NMO, which is still considered the most detailed clinical case study of its time [97]. A short time later, a case of bilateral optic neuritis and longitudinally extensive transverse myelitis (accompanied by involvement of the medulla oblongata) was reported by Herman Jakob Knapp (1832Knapp ( -1911 [305]. Knapp had been encouraged by Robert Bunsen (1811-1899), the father of spectral analysis, to study medicine under Hermann Ludwig Ferdinand von Helmholtz (1821-1894) in Heidelberg and had been made professor of ophthalmology in 1860. Later, in 1896, Theodor von Leber's doctoral student Karl Katz published his dissertation on NMO in Heidelberg [136]. Leber, eponym of Leber's hereditary optic neuropathy, a former pupil of Helmholtz and assistant of Knapp, had been made head of the Heidelberg department of ophthalmology in 1890. Nowadays, the Neuroimmunology Group at the University Hospital Heidelberg attempts to honour that heritage by dedicating much of its work to improving the diagnosis and treatment of NMO as well as deepening our understanding of the pathogenesis of this rare syndrome (e.g., [1,3,4,10,49,[51][52][53][306][307][308][309]).

Conclusion
The recent discovery of AQP4-IgG in patients with NMO has revived interest in the co-occurrence of eye disorders and myelitis. We believe the time has come to do justice to those who were the first to report on this rare yet intriguing syndrome by acknowledging their achievements and underlining their position in the history of medicine. The reports on amaurosis in patients with spinal cord inflammation by Ware, Baillie, Sichel, Carron du Villard and Pétrequin, re-presented here, are the earliest descriptions of possible NMO in the Western medical literature and thus deserve to be remembered. In addition, we point to the concept of 'spinal amaurosis', a long-forgotten precursor of the concept of NMO, and emphasise the important role played by ophthalmologists in first describing the syndrome. We hope that our findings may kindle further research into the early history of NMO.