be merely clothing in my own language facts with which you are all familiar, and in the advancement of which many of you have taken an important part. I further found that standing on the high pedestal on which you have placed me, I was, to a certain extent, placed above criticism, and therefore could not fairly deal with polemics.

It finally occurred to me that I might find some devious path or unbeaten track in the vast field of medicine which I might pursue with some measure of success. I intuitively turned my attention to the circulation, the ramifications of which pervade the whole field, and as writers hitherto on this subject have almost invariably traced the circulation from the centre to the periphery, it occurred to me that we might get a fresh view if we turned our attention in the opposite direction. There are numerous treatises on diseases of the heart and aorta, but until recent years a careful study of the peripheral circulation has been largely left to physiologists and pathologists. The experimental work of Cohnheim will ever remain a landmark in the pathology of the circulation, while to the school of Ludwig physiologists are no less indebted. To physiology medicine owes much, and all great advances are being prosecuted along physiological lines. If there have been any apparent divorce between the scientific basis and the practical application of our art, it is not due to any too rapid advance of physiology, but to physicians being too slow to fructify the field which has been tilled by physiologists. I have previously asserted that diseases of the heart most frequently arise from causes acting on the periphery, and hence there is here no room for specialism. The man who only studies the circulation with the aid of a stethoscope is a positive danger to society. I can, therefore, with an easy conscience and a sense of much satisfaction, devote some attention to that periphery.

The capillaries through which the interchange of nutritive pabulum and gases takes place between the blood and tissues, play a most important rôle in the animal economy. Yet they have received very inadequate atention from clinicians. Perhaps it has been thought that their structure and position could be so briefly described that any circumlocution in their description was unnecessary. But however simple their structure, and however apparent their functions, they constitute a vast filter bed for conveying nutritive material and oxygen to the tissues and for removing waste products therefrom. A careful study of how these changes take place, and how the functions of these little tubes are carried on, has always seemed to me a matter of as much importance as even the action of the heart itself. These little vessels are of extreme tenuity and delicacy, consisting of a single layer of endothelium, yet they are much stronger than most people imagine, and are capable of standing considerable internal pressure; they vary from about 0.5 to 1 millimetre in length, and from 7 to 13 micro-millimetres in diameter.

They are to a certain extent elastic, or at least they have the capacity of adapting themselves to the amount of blood which is driven through them. Their importance has been aptly described by Leonard Hill, who says: "The blood is brought into intimate relation with the tissues by diffusing through the endothelial wall of the capillaries, and this wall is of great tenuity; thereby takes place that change of material which maintains the combustion of the body and. the fire of life.

The capillary bed is a vast territory which pervades every tissue and organ of the body, and so numerous are these little vessels that it would be difficult to stick the point of a needle in any vascular area without wounding one or more, but in neurotic individuals you may wound many such vessels without drawing blood. In very plethoric individuals and in cases of polycythaemia, the capillaries of the body are fairly replete, but in ordinary mortals, especially in those of neurotic temperament, perhaps not a third of the capillaries are full at any one moment. Apply a sinapism to a very pallid skin, and you may wonder where all the turgid capillaries have sprung from. From the fact that under normal circumstances a sufficient quantity of blood cannot get through the arterioles to keep the enormous capillary bed full, the lateral pressure and the velocity in the capillaries are ever-varying quantities. When Leonard Hill stated that the pressure in the capillaries under certain conditions is often over 100 mm. of mercury, I thought that there must be some error of observation, as I was under the impression that such pressure would rupture these delicate little vessels, but I remembered the old advice: Do not think; try. I tried, and found that Leonard Hill had rather understated the fact, as I found variations from about 50 to 2,000 mm. of water. I also found equally great variations in the velocity of the blood in the capillaries. In text-books on physiology it is put down from 0.2 to 0.75 mm. per second; but my observations have given records from about 0.5 to 25 mm. per second. The capillary bed covers an enormous area; take, for example, the lungs, where all the air vesicles are surrounded by a meshwork of capillaries, and the surface of the air vesicles in the average individual has been calculated by Zuntz at 90 square metres. Numerous attempts have been made to estimate the capacity and sectional area of the capillaries, but in my opinion these questions are still unsolved. The method adopted of estimating the sectional area of the systematic capillaries is simplicity itself. We all know that with any given force the velocity is inversely as the sectional area. The mean velocity in the aorta has been set down as 320 mm., and in the capillaries as 0.5 mm. in the second; therefore, on this basis of calculation the sectional area of the systematic capillaries would be 640 times that of the aorta. It is not difficult to show that the premises are wrong, so it is highly improbable that the conclusion can be right. It at once becomes absurd if

The

we consider how this calculation would work out for capacity; if we reckon the average length of a capillary as 0.75 mm., and the length of the aorta as 480 mm., accordingly the aorta should hold as much as all the systematic capillaries. From the effective mean lateral pressure in the arteries it is difficult to draw any conclusion as to the velocity in the aorta, because the force imparted to the blood in the aorta by the heart is an ever-varying compound of kinetic and potential energy. Even if you did arrive at any fairly accurate idea as to the mean velocity in the aorta, it would not be correct to draw any inferences from a comparison between the velocity of the blood in the aorta and in the capillaries, because the conditions determining the velocities in the two sets of vessels are not comparable. higher the potential in the arteries the greater the velocity in the capillaries, but as this arterial potential is induced by obstruction to the outflow the velocity in the arteries will be diminished. As Leonard Hill appropriately says: "The circulation of the blood follows certain definite laws; unfortunately, the conditions of the flow are so complicated that these laws remain for the most part undetermined. A viscous fluid driven by an intermittent pump, which circulates through a system of branching elastic tubes of varying capacity; a system of tubes into and out of which passage of fluid takes place either by osmosis, filtration, or secretion; a fluid which varies in viscosity, a pump which varies in force, and tubes which have an ever-changing diameter and co-efficient of elasticity."

In a paper on tubal nephritis published in 1883, and in one on the pathology and treatment of dropsy in 1886, I dealt with the capillary circulation. I have long been in the habit of estimating the velocity by compressing the blood out of the capillaries in a given area and then watching the quickness or velocity of the return. This has served, and still serves, my purpose, but when I wish to record my observations I use a glass rod 10 millimetres in diameter. With the flat end of this rod I compress the capillaries, and then with a stop-watch recording fifths of a second I time the period of the return of the blood. If you divide the radius of this rod (5) millimetres) by the time, you get the velocity per second. For these observations you must select some spot where there is a network of capillaries which you can completely empty, such as those in the back of the hand or finger, and you must also choose a spot where the return current flows from all parts of the circumference. This method is so simple and accurate that it is a matter of surprise to me that, so far as I know, it has never been thought of before.*

*Dr. George Oliver has drawn my attention to a capillary dynamometer devised by Dr. Alexander Haig for gauging the amount of uric acid in the circulation and estimating the blood pressure. Dr. Haig compresses the blood out of the capillaries under pressures varying from about 5 to 20 oz., and times the periods of compression and of the capillary reflex. Dr. Haig says that his "instrument gives a constant definite area of pressure-a definite and measurable force, the pressure being applied for a definite and constant time, measured

When fluid is circulating in a capillary tube, the axial velocity is double the mean velocity. New, the erythrocytes travel in the axis, but as they occupy at least four-fifths of the lumen of the vessel, the mean must be fully 0.8 of the observed velocity. For estimating the pressure in the capillaries I use a modification of v. Kries's appar atus. For applying the pressure I use three sizes of glass plates measuring 20, 100 and 400 sq. mm., so a gram pressure on each of these plates represents respectively 50, 10, and 2.5 mm. of H2O. As before stated, I have recorded capillary pressures varying from 50 to 2,000 mm. of water, and my velocity records have ranged from about 0.5 mm. to over 25 mm. per second. Any one with a capillary velocity at the level of the heart which physiologists set down as normal might appropriately take up the refrain, "The hour of my departure's come."

The study of the lateral pressure and velocity of the blood in the capillaries is an exceedingly interesting one. A combination of these two forces represents the energy of the blood in the capillaries, and no doubt this energy is derived from the heart, and stands in direct relationship to the force of the cardiac contraction; the greater the force of the cardiac output the greater will be the energy in the capillaries, but the component elements of this energy-lateral pressure and velocity-need not bear any direct relationship to those respective elements in the arteries. These two conditions (velocity and pressure) might be said to stand, within certain limits, in an inverse ratio to one another, the more rapid the flow the less the lateral pressure, and vice versa. The lateral pressure depends on the statical condition of the blood, and just in proportion as you introduce movement you convert the force of pressure into that of velocity.

If you wish to drive a certain quantity of fluid through a tube, the velocity will depend on the force of the propulsion minus the obstruction to the outflow, with the inertia or viscosity of the fluid. (there is no fluid perfectly mobile) and the friction of the tube; and the lateral pressure will increase as the outflow is obstructed-in short, as the statical condition is maintained. The vis viva or energy of the blood in the capillaries can be répresented, as in any other vessel, by the formula The component forces of this energypressure and velocity-are constantly varying, so, for the sake of clearness, it will perhaps be better to describe them separately.

The pressure stands in direct relation to the freedom of the inby a metronome beating half seconds, the length of time the blood and color take to return being measured by the same instrument." Our methods are similar, but our objects are different. At one time I thought of drawing up rules for estimating the arterial blood pressure by the capillary velocity, but I soon found that the necessary corrections on account of the contraction or dilation of the arterioles, the position and temperature of the part under examination, and perhaps the viscosity of the blood were so numerous as to materially lessen the clinical value of any such method.

flow and the obstruction to the outflow. For example, take a very cold hand: the arterioles and small arteries may be so contracted that the mass of blood supplied to the capillaries is greatly diminished, and the lateral pressure correspondingly falls. Even in the arterioles there may be such a drop in the pressure-gradient that there may be a difference of 50 mm. of Hg between the lateral pressure in the digital artery and that in the radial. In cases of local syncope the lumen of the arterioles supplying the affected district is obliterated and the capillary pressure is reduced to nil. On the other hand, if you warm the hand, or take a glass of whisky, which dilates the arterioles, the mass of blood in the capillaries is augmented and the pressure rises; and the fall in the pressure-gradient between the arteries and capillaries becomes more gradual. An increased obstruction in the arterioles over a wide tract, such as the splanchnic area, raises the general arterial pressure and lowers the capillary pressure in the area supplied by the contracted arterioles.

As Cohnheim long ago pointed out, if you obstruct the outflow by tying a ligature around the limb, you greatly raise the pressure in the veins and capillaries distal to the ligature, but as you cannot thus completely obstruct the venous return without at the same time obliterating the arterial supply, the pressure in the veins does not rise so high as that in the capillaries, and the pressure in the capillaries does not attain the level of that in the arteries, and, of course, that in the obstructed artery does not rise above the general arterial pressure at the same level. With any given energy in an artery the pressure and velocity in the capillaries supplied by that artery stand in an inverse ratio to one another; the greater the pressure the less the velocity.

I have corroborated v. Kries's observations as to the effects of gravity on the capillary pressure, and like him I have found that the increase is usually less than one-half the hydrostatic effect-for example, if you take the capillary pressure in the finger at the level of the vertex, and then take it when the finger is lowered, say 600 mm., the increase may be only 200 mm. of water in place of an increase of 600 mm. of blood which it would be in an artery. I have also found that the increase is not at all uniform. It may vary enormously in different individuals, and in the same individual under different conditions. It largely depends on the condition of the vasomotor mechanism of the part which you are examining. If the arterioles be contracted and the inflow to the capillaries be obstructed, the increase may not be a third of the hydrostatic pressure, but if the arterioles be much dilated the increase may be half or even two-thirds of the increase which has taken place in the artery. Leonard Hill has pointed out to me that this increase chiefly takes place when the limb is immobilized; and when active movements are going on the blood is compressed out of the capillaries and this