It is evidently impossible that I should attempt, within the compass of a single lecture, any detailed examination of the very numerous cases reported in the Homoeopathic Treatises and Journals.  Having been in the habit of receiving the French “Archives of Homoeopathic Medicine” until the premature decease of that Journal, I have had the opportunity of becoming acquainted somewhat with the style of these documents, and experiencing whatever degree of conviction they were calculated to produce.  Although of course I do not wish any value to be assumed for my opinion, such as it is, I consider that you are entitled to hear it.  So far, then, as I am acquainted with the general character of the cases reported by the Homoeopathic physicians, they would for the most part be considered as wholly undeserving a place in any English, French, or American periodical of high standing, if, instead of favoring the doctrine they were intended to support, they were brought forward to prove the efficacy of any common remedy administered by any common practitioner.  There are occasional exceptions to this remark; but the general truth of it is rendered probable by the fact that these cases are always, or almost always, written with the single object of showing the efficacy of the medicine used, or the skill of the practitioner, and it is recognized as a general rule that such cases deserve very little confidence.  Yet they may sound well enough, one at a time, to those who are not fully aware of the fallacies of medical evidence.  Let me state a case in illustration.  Nobody doubts that some patients recover under every form of practice.  Probably all are willing to allow that a large majority, for instance, ninety in a hundred, of such cases as a physician is called to in daily practice, would recover, sooner or later, with more or less difficulty, provided nothing were done to interfere seriously with the efforts of nature.

Suppose, then, a physician who has a hundred patients prescribes to each of them pills made of some entirely inert substance, as starch, for instance.  Ninety of them get well, or if he chooses to use such language, he cures ninety of them.  It is evident, according to the doctrine of chances, that there must be a considerable number of coincidences between the relief of the patient and the administration of the remedy.  It is altogether probable that there will happen two or three very striking coincidences out of the whole ninety cases, in which it would seem evident that the medicine produced the relief, though it had, as we assumed, nothing to do with it.  Now suppose that the physician publishes these cases, will they not have a plausible appearance of proving that which, as we granted at the outset, was entirely false?  Suppose that instead of pills of starch he employs microscopic sugarplums, with the five’ million billion trillionth part of a suspicion of aconite or pulsatilla, and then publishes his successful cases, through the leaden