The structure and functions of the cerebellum examined by a new method. By Sir Victor Horsley, FRS, FRCS and R.H. Clarke, MA, MB. Brain 1908: 31; 45–124.
Robert Clarke and Victor Horsley seem an unlikely pair: the former a sportsman being a good judge of a horse, a keen huntsman, a first-rate shot, excellent cricketer and golfer, and a raconteur mainly of ghost-stories whose scientific interests were equally wide-ranging; the latter a pioneer of neurological surgery whose later life was increasingly distracted by controversial excursions into suffragette politics, the defence of vivisection, and evangelical patronage of the National Temperance League; but they shared an interest in gadgets.Horsley and Clarke start by debating whether or not the cerebellar cortex is directly connected to the peduncles and spinal cord. Marchi had described a descending cerebello-spinal pathway: Ramon y Cajal referred ambiguously to a via descendente and illustrated fibres in the superior cerebellar peduncle derived from the cortex; Probst correlated fibre loss in the peduncles with the amount of damage to cerebellar nuclei; and that no such connections existed was the opinion of Ferrier and Turner, and Risien Russell and Thomas. But none of the work on which these conclusions were based could be relied upon: what is needed in order to demonstrate the true anatomy of the cerebellum—and, pari passu, its function—are discrete but severe lesions of the cortex, sparing the nuclei, and with the course and destination of fibre degenerations traced using Marchi's method. In the end, the answer is clear: if the cerebellar nuclei are spared—in monkeys, dogs and cats—no fibres degenerate in the peduncles or spinal cord. Thus, the cerebellar cortex receives afferent fibres but its outflow is to the adjacent folia and the cerebellar nuclei. Now it seems desirable to study the effects of damage to the nuclei themselves. These are small, deeply situated and relatively inaccessible; so how can discrete and accurate lesions be made? Neither puncture with a small knife, galvano-cautery, nor the injection of acids or other fluids will do, and destruction by sparks or Faradism proves unsatisfactory. However, electrolytic lesions delivered by an insulated needle seem altogether more promising, yielding precisely defined necrotic lesions that fade off abruptly into undisturbed tissue. All that remains is to devise a method for accurate placement of these lesions at sites of interest.
In developing a stereotaxic instrument, exact relationships between landmarks on the outside of the head and its encephalic contents must be known. A first step in the right direction is the myelotom of Wilhelm Trendelenburg in which a model of the sagittal section of the brain is fixed above the head and a steel wire knife manipulated around that 2D model in the vertical plane by a sliding upright—these movements being reproduced in the brain by a ‘lazy-tongs’ arrangement. Although representing an advance on the cannulae and hook-like stillettes devised by Veyssière, this is not entirely satisfactory. Better is the method of rectolinear topography and stereotaxic instrumentation that directs an insulated needle to any point in the brain for stimulation or electrolysis. But how to obtain reliable coordinates of an amorphous structure (the brain) encased in an irregular sphere (the head) separated by muscle and a mobile integument (the scalp) is not so easy. Step one is to segment the contents of the cranium into eight parts (right and left frontal, occipital, temporal and cerebellar) by ‘slicing’ it in three planes—sagittal, horizontal (transverse) and frontal (coronal). It follows that each block has three flat internal surfaces placed at right angles to each other from which points of interest can be approached using perpendiculars from these boundaries. In this way reference to the outer irregular surface of each ‘cube’ is not needed. Next are the challenges of relating points of interest within each block to the perpendiculars dropped from these planar surfaces, and how to exteriorize the contours of these virtual segments for stereotaxic exploration of the living brain: ‘a selected point must therefore be known from the measurement of other heads, and can be trusted only so far as these data are constant’.
Horsley and Clarke start by taking a line connecting the lower margin of the orbit and the external auditory meatus as their horizontal reference; the frontal (coronal) line passes through both auditory meati; and the saggital line bisects the head between these two. A few inter-species adjustments are needed for placement of the horizontal line in order to ensure that this does indeed refer to corresponding parts of the encephalon in Macacus rhesus, cats and hedgehogs. Next, they subdivide these blocks into cubic millimetre grids so that every internal locus is a known distance from each of the three internal planar surfaces (Fig. 1). A nomenclature is proposed: amongst the 2 000 000 mm3 of the Rhesus brain, ‘left frontal segment, lamella v. J. 6 refers to a cubic millimetre in the left frontal segment 5 mm to the left of the median sagittal plane, 10 mm above the horizontal plane and 6 mm in advance of the inter-aural or frontal (coronal) plane’. There follows a detailed account of how the internal contours of the sections of frozen heads of many sizes and from many animals are used to build up a profile of anatomical arrangements for each species. The main emphasis is on Macacus rhesus. Details are given of the devices used to plug the external auditory meati in order to fix the posterior reference point of the horizontal line, the drilling of each segment with ivory knitting needles to create the grids, and the brain saws used to section formalin fixed tissue—frozen to perfection in order to be neither too brittle nor too soft and hence prone to tearing.
|
‘The application of the foregoing facts to our experimental investigations [of excitation and electrolysis] has been effected by an instrument the general plan of which will be most easily obtained by an examination of the illustrations. It would be very tedious to follow a minute description of every screw and detail ...’. In short, the frame is screwed onto the skull orientated along the horizontal plane defined by the earplugs and lower orbital margin (Fig. 2A and B), and the sagittal and frontal (coronal) planes easily created by erecting a perpendicular and bisecting that line, respectively. A variety of adjustments can be made using a vertical sliding motion to raise the horizontal plane from the orbital-aural baseline in order to orientate this horizontal with internal brain structures when working with monkeys, geese, ducks, hedgehogs and cats (dogs are unsuitable for cranio-encephalic topography because their head sizes vary too much and the nerve tracts are poorly ‘marked’). Despite seeking consistency of anatomical arrangements, the variation of head size within each species occasionally requires further fine adjustments. Using transverse guides that allow movement across the surfaces of the stereotaxic instrument, these coordinates are then used ‘to direct a needle to any depth perpendicular to any section plane and at any distance from the other two, or, in other words, to any point of known distance from the three inner surfaces of any of the segments into which the head is divided by the three section planes’. The needle is an irido-platinum wire insulated close to its point by a glass capillary tube inserted though a sheath and known to have reached its desired location down to a depth of 40 mm, vertically from above or horizontally from behind and with transverse movement in the frontal (coronal) plane by external gradations. Technical aspects of manufacturing these needles and their glass sleeves, made by Mr Ritterhaus of Huntley Street, Tottenham Court Road, have been solved by Professor Jackson of Kings College.
|
Horsley and Clarke now rehearse the history of electrolytic lesions dating from the work of Sir Humphrey Davy (1809); the first such lesions created in the central nervous system by Sellier and Verger (1898); and the placement of thalamic lesions in the cat, dog and monkey by Gustav Roussy (1907). But the true pioneer seems to be Dr Golsinger reporting his technique in which a 20–40 mA shock yields 8 coulombs to produce a pin-sized lesion and 36 coulombs for one the size of a cherry in the brains of six dogs, respectively, to the Conférence des Médecins de la Clinique des Maladies Nerveux and Mentales (St Petersburg, 1895). For Horsley and Clarke, sparks at a frequency of 100/s lack control in the production of lesions; whereas, despite some chemical dissociation and gaseous mechanical distension of the tissue being injured, the anode of constant current produces lesions that can be controlled down to small groups of cells or bundles of nerve fibres depending on the size, outline and position of the electrode and the amount of current delivered. Two milliamperes produces about 1 mm damage for every 1 min of exposure—the lesion having a cavity, penumbral necrosis and surrounding oedema (Fig. 3) beyond which tissue is structurally and functionally intact: ‘. physiological proof . is that when, during an excitation experiment, a restricted electrolytic lesion has been made to mark some spot from which a definite response has been obtained, we have found, on advancing the needle another millimetre, that the uninjured tissue immediately adjoining is normally excitable’. Fresh lesions are characterized by loss of all tissue elements—glia, neurons, axons and blood vessels, perhaps with some sparing of myelin sheaths; and by 3 weeks, this material is undergoing phagocytosis (Fig. 4).
|
|
The second purpose of Horsley and Clarke's stereotaxic apparatus is to stimulate specific areas of the living brain—building on their own work and that of Ferrier, Hitzig, Nothnagel and Dupuy, Sherrington, Rijnberk, Lewandowsky, and Pruss and Lourié. First, they deal with the much debated issue of whether, by comparison with the cerebral cortex, the surface of the cerebellum is relatively inexcitable. They conclude that this interpretation is correct. Previous contrary opinions probably reflect spread of the stimulus to neighbouring and more excitable structures—their preference being for indirect stimulation of the spinal accessory nerve, the functional anatomy of which they explore in some detail. Rather, motor responses depend on stimulation of the highly excitable cerebellar nuclei. That faradic stimulation produces a movement, say of flexion, and galvanic excitation the reciprocal extensor movement is explained by the principles of complex representation and reciprocal innervation already described by Hughlings Jackson and Sherrington, respectively. The cerebellar cortex is exclusively a sensory organ.
In describing this stereotaxic apparatus, Horsley and Clarke lean on their own previous account of efferent connections of the cerebellar cortex exclusively to its nuclei; and they advertise a further paper that is to follow describing functional properties of the cerebellum. Sadly, this did not appear in the 8 years prior to Horsley's death in Mesopotamia from heat-stroke on July 16, 1916; although Sir Victor did publish a paper in Brain (1909) with (Sir) Arthur MacNalty on the anatomy of spino-cerebellar connections. Nor, having settled the issue of no direct efferent projection, do Horsley and Clarke really deal with the problem that stimulated this work—tracing connections between the cerebellum and the spinal cord. Their main concern is the nuts and bolts of the apparatus made by Messrs. Swift and Son, Tottenham Court Road at a cost of £300. Nevertheless, ‘The structure and function of the cerebellum examined by a new method’ has since achieved iconic status, recognized by the imprimatur of Garrison and Morton's Medical Bibliography (item 4879.1) as having opened the way to stereotactic surgery of the brain—a method for studying and treating disorders of the central nervous system that remains active in contemporary clinical neuroscience, as exemplified by the papers from Alessando Stefani and co-workers (page 1596) and Jan Herzog and colleagues (page 1608) in the current issue.
Cambridge
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||