In the second place, the correctness of the conclusion must very greatly depend on our assigning the true value to each portion of the evidence, especially if the group of symptoms be a very com plex one. We still form our judgment from the aggregate, but we know that one part is much more trustworthy and more im portant than the remainder. One single symptom even may, by Its presence or absence, turn the balance of evidence in favor of one disease, or exclude another; but this view of ita importance in connection with the whole group, of which it is but a part, is very different from the error already pointed out of regarding any sign as " pathognomonic." On this point correct general know ledge of disease can alone give precision to our judgment ; but it is also the province of a work -on diagnosis to assign in some measure to each symptom its relative value. .
In the third place, the verification of the result wholly depends upon the accuracy of our knowledge of the theory of disease. The evidence of symptoms properly arranged leads us so far in the right direction for discovering its true seat and nature ; but it does no more than point out a number of requirements with re ference to particular organs, or to the system at large, which any disease must be known d priori to fulfil, before we can admit it to be that which exists in the case before us.
From these considerations, I think it must be evident that the more numerous and the more simple the symptoms are on which we have to decide, the more certain must be our diagnosis.
Further illustration may perhaps be deemed unnecessary, but my meaning may be made more evident by comparing the inves tigation of a case to the properties of figures in geometry. Sup pose that through any four fixed points straight lines are drawn enclosing a quadrangular space ; it is manifest that the number and variety of figures which may be produced is very great ; and if these figures are placed side by side and compared with each other, they will only be recognized as being four-sided figures, and few persons could find out that they had any other property in common. But if through two of the points (the first and third, for instance) the lines are always drawn parallel to each other, the number of instances is at once much reduced, and this fact is im mediately recognized as being common to them all. If, in addi tion to this,the lines drawn through the second and fourth points are also parallel, the class becomes reduced to those known as parallelograms, of which the opposite sides and angles are equal, and the original property of their passing through four fixed points becomes much more apparent. Further, if one of the an
gles is ascertained to be a right angle, we are certain that only one figure can fulfil all these several indications.
Again, the parallelograms may be compared with each other by the relative length of their diagonals, and we find that in the rectangular parallelogram the two diagonals are exactly equal. Here, then, we may disregard all the other facts, and finding straight lines drawn through four fixed points, inclosing a quad rangular space of which the diagonals are equal, we are certain that the opposite lines are equal and parallel, that all the angles are right angles, and that only one possible figure can possess these two properties, just as before we found that only one figure could possess all the other properties in detail. At the same time, if any one of these properties could not be detected on further investigation, we know that we must have made some mistake in the observation regarding the equality of the diagonals.
In studying disease, it is manifest that attention to one system only cannot lead to truth, since the causes of its production may be various ; but when a greater number are considered, and are found to harmonize together, the possibility of the whole group being produced by one or other of several causes becomes neces sarily very greatly diminished. When the symptoms present are obscure or uncertain, it is much more difficult to trace them back to their true source than when they are clear and intelligible. But yet we must remember that even after we seem to have arrived at a correct result from the comparison of two or more definite symptoms, yet if other important phenomena which ought to be found on closer search are absent, we must have committed an error in observation, and the opinion formed ought only to be persisted in when this exact correspondence can be traced, or good reasons can be assigned for the existence of an exception. Hence it sometimes happens that future examination of the same case, by bringing to light new symptoms, may oblige us to dis card an hypothesis framed on insufficient premises : indeed, we must often suspend our judgment altogether, till the progress of the' case has determined the actual form which the disease is about to assume.