remarkable examples of it. There is, indeed, scarcely any doctrine which may not in the hands of an ingenious person be wrought up in this manner into a fair system, amply supported by facts; and it is obvious that nothing can be more contrary to the rules of sound inquiry. On this ground, we may even make it a rule to receive with suspicion the statements of a writer, when we find him first proposing his doctrine, and then proceeding to collect from all quarters facts in support of it. Such a mode of investigation must be considered as contrary to the principles of fair induction; for these lead us first to take a full view of the facts, and then to trace the principles or doctrines which arise out of them.

IV. Receiving as facts on which important conclusions are to be founded circumstances which are trivial, incidental, or foreign to the subject. For example, in the investigation of affections of the spinal cord, appearances have been often considered as indicative of disease, which we have good reason to believe have arisen merely from the position of the body after death. In the same manner, in the investigation of a certain class of diseases, an important place has been assigned to slight appearances in the gastro-intestinal membrane, which, we have reason to believe, are entirely incidental, and worthy of no confidence in a pathological inquiry

V. Receiving as facts statements which falsely obtain that name. The sources of fallacy to be kept in view under this head are chiefly the following:

(1.) Receiving as facts statements which are not facts but opinions.-A person dies after being affected with a certain set of symptoms, and we find, on examination after death, the usual appearances of hydrocephalus. Another is seized with similar symp toms, and recovers. He is therefore said to have

recovered from hydrocephalus, and such a statement is often given as a medical fact. The man's recovery from certain symptoms is a fact; that he recovered from hydrocephalus is not a fact, but an opinion.

(2.) Receiving as a fact a statement which only assumes the relation of facts.-A person recovers from a particular disease, while he was using a particular remedy. His recovery is ascribed to the effect of the remedy; and the cure of the disease by this remedy is often given as a medical fact. The man's recovery is a fact; and that he used the remedy is another fact; but the connexion of the remedy with his recovery we are not entitled to assume as a fact: It is tracing between the facts the relation of cause and effect, a process of the utmost delicacy, and not to be admitted on any occasion without the greatest caution.

(3.) Receiving as facts general statements, or the generalization of facts. One of the most common examples of this error occurs, when a statement is given of a symptom or set of symptoms as certainly diagnostic of any particular disease, or of a particular morbid condition of an internal organ. Such a statement we hold to be of no value, unless we have absolute confidence in the narrator, both in regard to his habits as a philosophical observer, and to the extent of the observations on which his statement is founded. But, with every possible advantage in these respects, we are to exercise the utmost caution before we receive the relation thus stated as a fact; for it is to be kept in mind, that it is not properly a fact, but a generalization of facts. Some writers, for instance, have maintained with much confidence that a particular state of rigidity of some of the limbs is distinctly characteristic of ramollissement of the brain. But farther observation has shown that the disease may exist without this symptom, and that this condition of the limbs may appear in connexion with other diseases. Their observation

of facts was in so far correct, that this state of limbs does very often accompany ramollissement of the brain; the error consisted in giving it as a general fact, or a fact applicable to all cases of ramollissement-which is without foundation. Yet such statements, when brought forward with confidence, are often received as facts, and rested upon as established principles; and then the facts by which their fallacy might be detected are apt to be overlooked or forgotten.

This may perhaps be considered as one of the most prevailing errors in the modern science of medicine; and it is indeed astonishing to observe the confidence with which such statements are brought forward, and the facility with which they are received as equivalent to facts, without attention to the manifold sources of fallacy with which they are encumbered. Does a writer, for example, tell us he has ascertained that the spinal cord is diseased in all cases of tetanus. If we knew that such a statement had been founded on the careful observation of a hundred cases, it would be of value; if it was deduced from a few, its value is greatly diminished. But even if it had been deduced from the larger number, certain doubts would still arise in considering the relation thus stated as a fact. We should naturally ask ourselves,-was the narrator qualified to judge of the facts and their relations,were the cases referred to all really cases of tetanus, -were the appearances in the cord such as could properly be considered as indicating disease, or might any of them have been mere changes of colour, or other incidental appearances, which might have taken place after death, or might have been the effect of the convulsion rather than its cause,-or were they such changes as may be found in other cases without any symptoms of tetanus? Other sources of fallacy will come into view, if the statement be, that the nar rator has uniformly found a certain remedy of great

efficacy in a particular disease. Here, in the first place, similar questions occur as in the former instance ;-on how many cases did he found his statement, how did he ascertain the disease,—and was he qualified to decide that it really was a case of the disease which he alleges? But, supposing all these questions to be answered in a satisfactory manner, others still arise, namely,--had the alleged treatnent really any influence on the recovery of the patients, did they get well in consequence of the treatment, or in spite of it, or altogether independently of it, have not similar cases recovered spontaneously, or under modes of treatment entirely different? Such is the uncertainty of causation and generalization in medicine; and such is the danger of receiving general statements as equivalent to facts.

VI. In forming a collection of facts on which we are to found any conclusions, it is always to be kept in mind that fallacy may arise from the absence of important facts, as well as from the reception of statements which are untrue. Hence the erroneous conclusions that may be deduced from statements which are strictly true; and hence the fallacious systems that are built up with every appearance of plausibility and truth, when facts are collected on one side of a question, or in support of a particular doctrine.

In forming a collection of facts, therefore, as the preliminary step in any inquiry, the following rules ought to be kept strictly and constantly in view before we advance to any conclusions:

I. That all the facts be fully ascertained,-that those collected by ourselves be derived from sufficient observation,-and that those which we receive from others be received only on the testimony of persons fully qualified to judge of their accuracy,

and who have had sufficient opportunities of acquiring them.

II. That the statement include full and fair view of all the facts which ought to be taken into the investigation; that none of them be disguised, or modified so as to be made to bear upon a particular doctrine; and that no essential facts be wanting.

III. That the statement do not include facts which are trivial, incidental, or foreign to the subject.

IV. That we do not receive as facts statements which are not facts, but opinions or general assumptions.

SECTION II.

OF ARRANGING, COMBINING, AND SEPARATING FACTS.

THE precautions now suggested appear to be those which it is necessary to keep in view in making a collection of facts respecting any subject under investigation. Our next step is to arrange the facts according to the characters in which they agree; to separate from the mass those which appear to be only fortuitous or occasional concomitants; and to place by themselves those which we have reason to consider as a uniform and legitimate series or sequence. This is the first step towards tracing the relations of the facts; and in every investigation it is a process of the utmost consequence. In the other departments of physical science this object is accomplished by means of experiments. These are so contrived as to bear distinctly upon particular points; and by the result of them we are enabled to separate