The bleeding from fresh skin punctures gives a further criterion of the quantity of blood, within certain limits, but is modified by changes in the coagulability of the blood. Anyone who has made frequent blood examinations will have observed that in this respect extraordinary variations occur. In some cases scarcely a drop of blood can be obtained, while in others the blood flows freely. One will not err in assuming in the former case a diminution of the quantity of the blood. The fulness of the peripheral vessels however is a sign of only relative value, for the amount of blood in the internal organs may be very different. The problem, how to estimate exactly, if possible mathematically, the quantity of blood in the body has always been recognised as important, and its solution would constitute a real advance. The methods which have so far been proposed for clinical purposes originate from Tarchanoff. He suggested that one may estimate the quantity of blood by comparing the numbers of the red blood corpuscles before and after copious sweating. Apart from various theoretical considerations this method is far too clumsy for practical purposes. Quincke has endeavoured to calculate the amount of blood in cases of blood transfusion for therapeutic purposes. From the number of red blood corpuscles of the patient before and after blood transfusion, the amount of blood transfused and the number of corpuscles it contains, by a simple mathematical formula the quantity of the blood of the patient can be estimated. But this method is only practicable in special cases and is open to several theoretical errors. First, it depends upon the relative number of red blood corpuscles in the blood; inasmuch as the transfusion of normal blood into normal blood, for example, would produce no alteration in the count. This consideration is enough to shew that this proceeding can only be used in special cases. It has indeed been found that an increase of the red corpuscles per cubic millimetre occurs in persons with a very small number of red corpuscles, who have been injected with normal blood. But it is very hazardous to try to estimate therefrom the volume of the pre-existing blood, since the act of transfusion undoubtedly is immediately followed by compensatory currents and alterations in the distribution of the blood. No property of the blood has been so exactly and frequently tested as the NUMBER OF RED CORPUSCLES PER CUBIC MILLIMETRE OF BLOOD. The convenience of the counting apparatus, and the apparently absolute measure of the result have ensured for the methods of enumeration an early clinical application.