_Note._ This apparatus was first described in the third part of Dr. Beddoes’s Considerations; its relation to Mr. Davy’s experiments with the improvements it has lately received, may probably be deemed sufficient to excuse the re-printing it.—The weight I. Fig. 2, having been omitted in the plate, the reader must supply the deficiency. W. C. PROPOSALS FOR THE PRESERVATION OF ACCIDENTAL OBSERVATIONS IN MEDICINE. In times beyond the reach of history, the medicinal application of substances could have arisen from no other source than accident. Among articles of the materia medica of known origin, we are indebted to accident for some of the most precious. Accident is every day presenting to different individuals the spectacle of phænomena, arising from uncommon quantities of drugs on the one hand, and on the other, from uncommon conditions of the system, where ordinary powers only have been knowingly or recently applied. What is said of drugs may be extended to natural agents and mental affections. From conversation with a variety both of medical practitioners and unprofessional observers, the author of this proposal is persuaded that such authentic occurrences only, as have presented themselves to persons now living would, if they could be brought together, compose a body of fact, so instructive to the philosopher, and useful to the physician, that he despairs of finding a term worthy to characterize it. In some cases, the influence of unsuspected powers would be detected. In others, resources available to the purpose of restoring health in desperate situations would be directly presented, or could be detected by a short and easy process of reasoning. Some anomalous observations, by shewing the absence or agency of contested causes, would perform the office of _experimenta crucis_—Unusual affections occur of which an exact account would be among the means of removing from physic its opprobrious uncertainty: for this uncertainty frequently depends upon our inability to distinguish the subtler differences in cases which resemble each other in their grosser features. No striking fact can be accurately stated, in conjunction with its antecedent and concomitant circumstances, without improving our acquaintance with human nature. Our acquisitions in this most important branch of knowledge, may be compared to a number of broken series, of which we have not always more than one or two members. But every new accession bids fair to fill up some deficiency; and a large supply would contribute towards connecting series apparently independent, and working up the whole into one grand all-comprehending chain.