On looking through some of the numerous observations on the treatment of melancholia by drugs, the most striking point is the diversity of opinion as to their use. This disagreement comes out perhaps most strongly in the case of opium, no doubt because it has been the medicine most widely tried. One leaves the study of the subject with the uncomfortable feeling that the whole drug treatment of melancholia is still quite empirical; for eminent observers press that opium should always be tried in every melancholic, while other no less eminent physicians are equally strong in condemning its use in every case. Our own excuse for adding to the over abundant literature on this subject is that we desire to point out some indication in the use of the two drugs, opium and sulphonal, to which we are convinced that more weight than is generally thought should be given. We have been repeatedly struck by the marvellous effects of opium in some cases of melancholia; and have as often been equally surprised at its utter failure in other cases closely resembling those in which we have been so successful. In hopes of finding some sufficient reason for such apparently contradictory results we have carefully examined the case books of this house for recent years, and have tabulated a sufficient number to warrant, we think, an examination of our conclusions by those who have wider opportunities. To these we have added a few non-asylum cases which we have been able to follow ourselves; which would go towards proving that these two drugs have an action independent of any advantage in any general asylum here. We do not propose to go fully into each case, as this would be unnecessary waste of space, and our remarks can be readily followed by a glance at the tables. No attempt has been made to classify, to arrange, or to select the cases beyond choosing only those in which opium or sulphonal has been given. One selection indeed we could not avoid. The material on which we have had to work being almost exclusively drawn from the upper and middle classes, it was inevitable that our patients should be drawn from the same classes; but this will probably not invalidate our conclusions.