malady under which an individual suffers carries with it certain pathological ideas, and conveys the case to its position in our nosological categories. But the examination of particular cases is the basis of all pathological science; observation of the phenomena of disease is the source whence our notions are derived; and hence the standard to which we must appeal for the confirmation or contradiction of our doctrines. Thus the classification we employ has its correctness put to the test at the very time that we make use of it; and to frame a satisfactory classification is one great object of natural science. In order that such classification should be absolutely perfect, our information should be co-extensive with the facts; and, until this point is reached, all systems are but temporary arrangements, which express only some portions of the truth of real natural affinity, and this by the formation of groups, the elements of which bear the greatest appreciable resemblance to one another.

It is not evident that any perfectly "natural system" of classification can exist the only natural division which is established beyond dispute is that of the individual. Such terms as variety, species, genus, order, class, sub-kingdom, and kingdom, have an unfixed meaning, and a changing application. They are not only convenient, but necessary aids in the progress of scientific investigation; their construction and usefulness are proofs and measures of the development of scientific knowledge; they convey meanings which, for practical purposes, are well understood: but their true meaning cannot be regarded as established; for high authorities differ widely as to the essential ideas conveyed by such terms as variety and species; and enlarged observation renders it necessary that either individuals should change their places, or that features regarded as indicative of specific difference should be held as characteristic only of variety, and vice versá; and lastly, that new nomenclature should sometimes be employed to convey accurately the truths we have ascertained.

The advantage of classification is the power of recognition

which it furnishes by the aid of certain features; the fault is that it cannot express all the features. The names by which we know things are sufficient for the former purpose; our information is insufficient for the latter. There is this hypothesis, moreover, underlying all our systems of classification; viz., that there exists between the individuals of a species, and between the species of a genus, some nearer, deeper, and more intimate relation than any which is apparent in petal, in organ, or in symptom; that in the ultimate essence-in that which makes the plant, animal, or disease just what it is, and no other—there is some resemblance to that of its congener, which may be witnessed to by external features, which determines the characters of those features, but which after all is not adequately expressed by them. For in one individual this character is absent, in a second that; and no one possesses all, and yet not more than all, the features proper to the group. There is an imaginary centre, an ideal type, around which all the members radiate, and from which they diverge. No two individuals are entirely alike; no two cases of what is called "the same disease" are precisely similar; and, on the other hand, no two natural objects are entirely different; they all possess some material or some properties in common. The same elementary substances, the same elemental forces, run through all Nature's works; their varied combinations produce the infinite variety of forms by which we are surrounded; and all our attempts to classify them result in the drawing of artificial lines, by which we may roughly separate from one another artificial groups, bring them into some relation with our laws of thought, study some of their characters, and compare them with those of other groups. We give names to the groups that we have formed; these names enable us to connote certain facts about them; and they convey to others, more or less accurately, the ideas we have formed of the relationship of their individual members: but names can neither register nor describe all the facts; and they fail to convey the whole of our ideas of the relation-which we can only obscurely see-between the objects they enable us to recognise;

for our knowledge is not in a position to allow of our assigning, with positive certainty, to the different objects that we see, the value of all the manifold materials, properties, and actions which they have in common.

But in this condition of our minds, and of their scientific information, no method is so useful as the numerical. It enables us to bring together those objects which present the largest number of features in common; to approximate the most nearly to a true valuation of particular characters,-those having the highest value which persist through the longest series of divergent elements;-to compare group with group; and to advance, by its means, to the separation from each other of those which agree in some few-they may be striking but inessentialqualities; and to detect the inner, most important, and most numerous relations among those which, in outer and conspicuous details, differ widely.

In the investigation of disease this method is of especial service. Minute observation, and its numerical analysis, compel us to rearrange some of our nomenclature, and in the case before us to remove from the group denominated epilepsy at least three forms of disease which have constantly passed under its name. Having accomplished this, we have a residue of cases presenting closer affinities than those which existed among the heterogeneous elements previously strung together; and the further examination of this group affords us some nearer approximation towards a knowledge of its natural history.

The value of a numerical result, in regard of pathology and therapeutics, is in proportion to the mutual proximity of the cases examined, and not simply to the number from which such result is derived. If our object is to know the natural history of convulsive diseases as a whole, we must of necessity embrace a vast number of cases; and some information gathered from such broad basis would possess great interest and utility; but, inasmuch as these diseases differ widely from one another, in their structural conditions, their phenomena beyond the convulsions, their influence upon social qualifications, upon

mind, upon life, and upon death, it is obvious that they must be regarded separately if we would learn the truth about them in these important relationships.

The discrepancies which occur between the statements of different authors,-writing with equal ability and with equal honesty, are due to the fact that they, in reality, deal with different diseases. The group of convulsive cases examined by one may be composed mainly of distinct cerebral diseases, such for instance as are presented by the epileptics of a lunatic asylum; that examined by another may have no lunatics among its number, but may be made up of eccentric convulsions or of diathetic diseases chiefly; whereas that described by a third may consist principally of epileptics. It is obvious, therefore, that the statistical results obtained from these three groups must differ widely in regard of such points as mental condition, hereditary taint, &c. Taking chronic convulsive disease as the sole ground of association for a number of cases; and forming several groups, upon that simple principle, in different countries, towns, or even in the same town, were that possible, the probability is that no two groups would contain the same proportion of the several different forms of convulsive affection. Nay, it is certain that they would not do so unless every group was so large that it embraced within itself all forms and conditions of variation. Statistical results are uniform only when the numbers from which they are derived are thus extensive. Until that point is reached there is variety, and, as it would seem, discrepancy; and especially has this been the case with regard to the subject now considered, for with so vast a concourse of diseases as those which make up the convulsive, such point can scarcely be reached, the objects be minutely examined, and the results compared, in the lifetime of any individual. What is required, then, is division of labour and of subject; and if this division is carried out as nearly as it can be in accordance with nature, and carried out to the furthest point that we can pursue it, a comparatively small number of cases will furnish some valuable information. For some purposes

it is better to have the statistical examination of eighty cases as nearly as possible identical, than that of eight hundred cases differing widely.

In the following volume I have therefore analysed eightyeight cases of simple and idiopathic epilepsy, this being the number of examples of that disease with regard to which I have been able to record accurate information. In them I could trace no evidence of any other affection, and to them I could apply no other name than epilepsy. The number, it may be said, is small, but the examples bear close resemblance to each other, and the result of their examination will, I think, be more valuable than if I had arranged in one large group for analysis all those cases of different chronic convulsive diseases—i. e. of so-called epilepsy-which have fallen under my observation and care.

If the principle which has guided me in applying the name epilepsy to them alone be held to be erroneous, it does not detract from the usefulness of these researches; for it is surely better that "idiopathic epilepsy" should be examined by itself, than that any conclusions with regard to its effects upon mind, upon motility, and upon the general health; or than that any statements as to the age of its onset, the frequency, form, or duration of its seizures, should be confounded by conclusions drawn from "symptomatic epilepsy," which includes every kind of lesion of the brain; and "sympathetic epilepsy," which includes diseases as numerous as the organs of the body, and as varied as the changes they may undergo.

Nothing, however, is more remote from my thought than that any of these results are final. Statistics do but deal with imperfect knowledge; they register so-called “ exceptions," which are either ignorances or errors; and they also point out the direction in which truth yet concealed may be discovered: but the object of scientific investigation is, not to rest satisfied with this recognition of an exception, but to find out its cause; not to stop when we have gained only some glimpse of the position of truth, but to advance, as far as possible, to the