NECESSITY FOR TRAINED BACTERIOLOGIST.--It is evident that the services of a trained bacteriologist are indispensable in the accurate diagnosis of plague, unless (as rarely is the case) the observer himself is both clinician and bacteriologist. Even in this case it is far better for two persons, clinician and bacteriologist, to work together. I will not discuss the technic of the procedures of biologic diagnosis, which is described by Dr. Schöbl in the preceding pages. Except under circumstances of necessity, the clinician should always turn this work over to the bacteriologist. Serum reactions, when present, occur too late to be of service in practical diagnosis. The necessary procedures of the biologic diagnosis include blood-culture, smear examination (microscopic) of aspirated material from the œdematous tissues surrounding gland masses and from glands themselves; examination of sputum smears and of thick-blood smears. All should be practised but, according to our Manila experiences, smear examinations of aspirated material and blood cultures are the most reliable methods, in the hands of a competent bacteriologist. Attention is invited to the reports of Dr. Otto Schöbl, already quoted. BACTERIOLOGIC PROCEDURE.--Dr. Schöbl was able to secure positive blood cultures, within 24 hours, from all of a long series of cases of plague, both bubonic and septicæmic. As much blood as it was possible to secure was aspirated from superficial veins and introduced into the culture media at the bedside, ten c.c. being secured whenever it was possible. The smear preparations for staining and culture inoculations upon slants were also made at the bedside from aspirated matter obtained from œdematous periglandular tissues or from gland puncture, an aspirating syringe being used. The drop or two of fluid which can be expelled from the hollow needle is usually sufficient for smears and tube inoculations. NON-BIOLOGIC DIAGNOSIS.--I do not contend that other diagnostic means than biologic ones should not be used in plague. On the contrary, it will inevitably happen at times that resort must be had to methods of diagnosis which are purely clinical. When this is the case, treatment, along lines to be detailed presently, should be instituted upon the establishment of a presumptive diagnosis. This presumptive diagnosis may be reached after due consideration of physical signs and symptoms. A carefully taken history of the onset and course of the disease will be valuable but unfortunately such histories can rarely be secured. It is far safer to mistakenly pronounce a case "plague" and to institute appropriate treatment, than it is to hesitate in the absence of a perfect clinical picture and to permit the golden moment for treatment to pass.