very different conclusions; which our readers will be better prepared to appreciate, if we first remind them of what careful anatomical research has within the last few years ascertained, as to the course of the bundles of nerve-fibres that are the continuations of the posterior roots, after their entrance into the spinal cord. None of the inquirers who have devoted themselves to this difficult investigation have done so much to elucidate the structure of the cord as Mr. Lockhart Clarke;* and what others have been able to demonstrate harmonizes fully with his results. Of those fibres of the posterior roots of either side which are brought into view by a transverse section, some appear directly continuous with the fibrous strands of the posterior column, and others with those of the lateral column of the same side; whilst others at once enter the posterior cornu of the grey substance of the cord. Of the latter some may be traced through the grey substance towards the anterior cornu of the same side; and appear to be continuous (whether they are so directly, or through the intervention of vesicular substance, is not yet certain) with motor fibres proceeding from it into the anterior roots; but the larger proportion seem to lose themselves in the grey substance. Of these it may be pretty certainly affirmed that some come into connexion with the cells of the vesicular matter; whilst others pass out into the lateral and posterior columns, a considerable proportion passing across the commissure, and, after passing through the grey substance of the other side, emerging into its posterior and lateral columns, some fibres appearing also to pass towards the anterior cornu of that side. When the course of the fibres of the posterior roots is brought into view by longitudinal sections of the spinal cord, it becomes apparent that some of the fibres which enter the posterior columns pass in the first instance upwards, and others downwards, in those columns; but that in neither case do they pursue this course far, their direction being soon changed to the transverse, so that they enter the posterior part of the grey substance a little above or a little below the fibres which passed into it horizontally from the nerve-roots. To the accuracy of Mr. Lockhart Clarke's statements on these points, we can bear testimony from our own examination of his preparations; and their very remarkable accordance with the conclusions drawn by M. Brown-Séquard from his experimental inquiries, is not a little confirmatory of our belief in the latter, especially when it is considered that the greater part of these results were obtained without any knowledge of what anatomical investigation could reveal, M. BrownSéquard's researches having been carried on contemporaneously with those of Mr. Lockhart Clarke.

In the attempt to determine experimentally what is the part of the spinal cord which serves to convey sensory impressions, it is of great importance to observe certain precautions, the neglect of which is very likely to lead to erroneous conclusions. The first thing to be guarded against, is the disturbance of the functions of the spinal cord which often results from the operation of laying it bare; diminution or even loss both of sensibility and of the power of voluntary motion in the • Philosophical Transactions, 1851 and 1858.

posterior limbs being so frequent a result of the opening of the spinal canal, that some physiologists regard it as inevitable. M. BrownSéquard, however, has found that the spinal cord may be laid open for a considerable part of its length, without any apparent diminution of sensibility or of power of motion (except such as results from the injury done to the muscles of the spine), if the operation be performed quickly, if pain be prevented by the exhibition of chloroform, and if any considerable loss of blood be avoided. It is of course essential that the persistence of the normal sensibility should be fully substantiated in the posterior extremities of an animal thus treated, before the effects of division of any part of its spinal cord are tested experimentally. Again, it is important to bear in mind that the powers of receiving and of conducting the impressions which give rise to sensations, do not by any means correspond with one another; so that a deficiency in the former is no disproof of the existence of the latter; or, in common language, a part may not be in itself sensible, although it may be the medium of transmission for the sensory impressions made on some other part. Of this fact the following curious example is given by M. Brown-Séquard:A large root of the sensory division of the fifth pair (the most impres sionable nerve in the body) passes down the medulla oblongata between the anterior pyramid and the corpus restiforme, towards the nib of the calamus scriptorius; it has been shown by Magendie that the division of this root by a transverse section of one half of the medulla oblongata occasions loss of sensibility of the face, so that this root obviously serves as the channel of sensory impressions; and yet if a large pin be inserted into the medulla oblongata in such a manner as to penetrate it, there is no sign of pain. Hence it is obvious that the absence of signs of pain, when any particular portions of the cord are irritated, does not justify the inference that those portions are destitute of the power of conducting sensory impressions. Thirdly, in making use of excitation, especially galvanism, as a means of determining the functions of the several parts of the cord, unless this be done in such a manner as to secure the limitation of the excitement, no trustworthy results can be attained. Now practically it is found to be extremely difficult, if not impossible, to effect this limitation; the application of galvanism, for instance, to either of the columns of the cord being almost certain to involve the contiguous roots, to an extent which there is no means of determining. And even if this difficulty can be excluded, the experimental use of such excitation must be extremely limited, since (as just shown) it only serves to indicate what parts of the cord are impressionable, and can give no valid information as to their relative conducting power.

Hence the method of experimenting by division of the several parts of the cord is the one which is for most purposes to be preferred; but it is requisite to use the greatest caution both in observing its results, and in drawing inferences from them. This method has to be put in practice in two modes: first, that part alone of the cord is divided whose functions we wish to determine, and the results of that division are tested; second, all the other parts of the cord are divided save that

one, and the results of that division also are observed. Now, of course, the results of these two operations should be antagonistic and at the same time harmonious; that they seem to be otherwise may be due, on the one hand, to an imperfect performance of the experiment, or, on the other, to an error in our observation or our interpretation of the phenomena. In making experiments of this kind, it is especially necessary to determine with precision the exact limits of the lesion inflicted. The most dexterous operator may find himself in error on this point, the section being often proved by subsequent examination to have been either more or less extensive than he had supposed it to be. And there can be little doubt that the want of such exact determination has been the cause of many of the contradictions between the results of operations apparently but not really the same. The more carefully this determination is made in each instance (by a subsequent examination of the spinal cord after it has been hardened by temporary maceration in alcohol), the more satisfactory do the results of sectionexperiments prove to be; since it is found that where the lesions are the same, their results are the same, and that perplexing differences in results of experiments supposed to be identical are explicable by differences in the lesions really inflicted.-In the care which he takes to avoid the foregoing and other sources of fallacy, we regard M. BrownSéquard as pre-eminent amongst the experimental physiologists who have devoted themselves to this investigation; and are disposed to accept his results as, if not containing the whole truth, the nearest approximations to it which have yet been obtained.

When both the posterior columns of the spinal cord are divided, but the remainder of the cord is left entire, not only is there no diminution of sensibility in the parts below the section, but there is an absolute increase, which is sometimes extremely remarkable. This result shows itself whether the section be made in the lumbar, dorsal, or cervical region; it is equally manifested after section of the restiform bodies, and in a less degree after a transverse incision in the cerebellum, in the processus a cerebello ad testes, or in the corpora quadrigemina. Putting aside for the present the phenomenon of hyperesthesia, and looking to the fact that there is no diminution in the sensibility of the parts below the section, it appears certain that the posterior columns are not essential to the conveyance of sensory impressions from the nerve-roots to the encephalon, but that those impressions can pass by some other channel. On the other hand, if the whole spinal cord except the posterior columns be divided, sensibility is lost in all the parts below the section except those at a short distance from it; though, if even only a very small part of the central grey substance remain undivided, a proportional measure of sensibility is still to be observed. The cause of the persistence of sensibility in the parts immediately below the section seems to lie in the fact, that the portion of the posterior roots which enters the posterior columns of the cord is not interrupted by the section of its remaining substance; but that the fibres of this portion do not proceed far in the posterior columns, seems proved by the circumstance that sensibility persists for only a short distance below the

section, the root-fibres of the parts further removed from it having already quitted the posterior columns for the grey substance, and the transmission of their sensory impressions being thus interrupted by the section of the latter. A result harmonizing with this is obtained from experiments of another kind. If the posterior columns alone be divided, indications of sensibility may be obtained by irritation of the upper cut surface, involving the nerve-roots which pass into it. These indications are not diminished by complete division of the same columns at some distance above; but if one section of these columns be made after another, nearer and nearer to the irritated surface, there is to be observed, first a diminution, and (as the distance is reduced) a gradual extinction of the sensibility; whilst, if a section of the remainder of the cord be made at some distance above the irritated section of the posterior columns, there is a like cessation of the indications of suffering upon the application of the irritation. From these facts it seems to be a fair inference, that some (at least) of the conductors of sensory impressions pass upwards for a short distance in the posterior columns; but that, instead of proceeding in the same course, they then quit those columns for some other part of the cord. But further, it was long since observed by M. Brown-Séquard that the lower or caudal surface of the section of the posterior columns is not less impressible than the upper or cephalic surface; and by pursuing the same method of experimenting, he has shown that this result is due to the passage downwards for a short distance, in the posterior columns, of a portion of the conductors of sensitive impressions; these conductors, however, soon altering their course, and passing first into the central substance of the cord, and then in it upwards or towards the brain.These results, it will now be seen, harmonize so remarkably with the facts determined by anatomical research, as to leave no reasonable doubt of the correctness of the conclusion,-that of the conductors of sensory impressions contained in the posterior roots of the spinal nerves, a part go upwards (or towards the encephalon) for a time in the posterior columns, and that another part bend downwards (or away from the encephalon) in the same columns; but that in each case, the sensory fibres soon quit the posterior columns to enter some other portion of the spinal cord.

We have now to inquire what can be safely inferred from experiment as to the relative share taken by the anterior columns, the lateral columns, and the central grey substance, in the transmission of sensory impressions to the encephalon. With regard to the lateral columns, the results seem to correspond with those obtained in experimenting upon the posterior; for whilst section of the lateral columns alone does not produce any diminution, but rather an increase, of sensibility in the parts below, section of all other parts than the lateral columns extinguishes the sensibility of all the parts below, save in the immediate neighbourhood of the section. Hence it seems, in the first place, to be a justifiable inference, that the lateral columns are not continuous conductors of sensory impressions from the nerve-roots to the encephalon; whilst in the second, it may be presumed, from the

persistence of sensibility in the parts a little below the section, that some of the fibres of the posterior roots pass along them for a time before entering the central substance; and this last inference is borne out by the results of experiments analogous to those just cited, which indicate that these conductors pass, as in the posterior columns, downwards as well as upwards, though for no great distance.-In regard to the anterior columns, however, the case is somewhat different; for whilst it is not found that section of them produces any notable reduction in the power of transmitting sensory impressions, it appears that some degree of that power remains when the entire spinal cord has been divided with the exception of the anterior columns. By Calmeil and Nonat it has been thought that these columns have a large share in this function; but according to M. Brown-Séquard, it is only when some portion of the central grey substance has remained undivided, that sensibility manifests itself immediately or shortly after the operation; the sensibility being, in his experience, altogether extinguished for a time, when the division of the central grey substance has been complete, and only reappearing in a very imperfect degree after the lapse of some hours.-If, then, neither the posterior nor the lateral columns serve for the transmission of sensory impressions from the nerve-roots to the encephalon, and the anterior columns have but a slight participation in this function, it would seem to be established, by the method of exclusion, that the central grey substance must be the principal channel for such transmission. We must own, however, that this inference does not seem to us so free from objection, as to be entitled to claim for itself acceptance without more support from experiment than we find adduced in M. Brown-Séquard's lectures. To make the proof complete, it should be shown, in the first place, that section of the grey substance alone, without the division of either the anterior, the lateral, or the posterior columns, almost entirely annihilates the sensibility of the parts behind the section; and conversely, that section of all those columns at once, the grey substance alone being left, has little effect in diminishing the sensibility of the parts behind. The first kind of proof it might not be possible completely to attain, since the whole of the grey substance of the cord could scarcely be divided without carrying the section through one or other of the columns; yet still some nearer approximation toit might be effected than we gather from M. Brown-Séquard's statements that he has made; and some more satisfactory explanation seems to us to be required, of the results obtained by Magendie, Sarlandière, Longet, and others, who found that the passage of a stylet down the central parts of the cord does not produce any marked effect either on sensibility or on power of movement. The second is surely attainable; and we gather from experiments cited elsewhere, that M. Brown-Séquard has firmly established the fact that the continuity of even a small part of the grey matter, the whole remainder of the cord being divided by transverse section, serves for the transmission of sensory impressions. Still this evidence by no means satisfies us that the grey matter is the channel through which these impressions are conducted all the way to the encephalon ;