the distance to be traversed by the blood while fulfilling these functions; and explain the importance of the comparatively slow rate at which it will be found to move through that short distance. The histological study of a typical capillary (see Fig. 10) shows that its thin wall is composed of a single layer only of living flat endothelial cells set edge to edge in close contact; and that the edges of the cells are united by a small quantity of the so-called cement-substance. If the capillary be traced in either anatomical direction, the wall of the vessel is seen to become less thin and more complex, till it merges into that of a typical arteriole or venule, the walls of which are still delicate, though less so than that of a capillary. That the capillary walls are so thin and soft, and are made of living cells, are very important facts as regards the relations between blood and tissue. It is of great importance for the variation of the blood-supply to a part that they are also distensible, elastic, and possibly contractile.

Direct Observation of the Flow in the Small Vessels.-The capillary flow is visible under the compound microscope, best by transmitted light, in the transparent parts of both warm-blooded and cold-blooded animals. It is important that the phenomena observed in the latter should be compared with observations upon the higher animals; but the fundamental facts can be most. fruitfully studied in the frog, tadpole, or fish, inasmuch as no special arrangements are needed to maintain the temperature of the exposed parts of these animals. Moreover, their large oval and nucleated red blood-corpuscles are well fitted to indicate the forces to which they are subjected. The capillary movement, therefore, will be described as seen in the frog; it being understood that the phenomena are similar in the other vertebrates. In the frog the movement may be studied in the lung, the mesentery, the urinary bladder, the tongue, or the web between the toes. During such study the proper wall of the living capillary is hardly to be seen, but only the line on each side which marks the profile of its cavity. Even the proper walls of the transparent arterioles and venules are but vaguely indicated. The plasma of the blood, too, has so nearly the same index of refraction as the tissues, that it remains invisible. It is only the red corpuscles and leucocytes that are conspicuous; and when one speaks of seeing the blood in motion, he means, strictly speaking, that he sees the moving corpuscles, and can make out the calibre of the vessels in which they move. The observer uses as low a power of the microscope as will suffice, and takes first a general survey of the minute arteries, veins, and capillaries of the part he is studying, noting their form, size, and connections. In the arteries and veins he sees that the size of the vessels is ample in comparison with that of the corpuscles; that, in the veins, the movement of the blood is steady, but in the arteries accelerated and retarded, with a rhythm corresponding to that of the heart's beat. In some parts, if the circumstances of the observation have somewhat retarded the circulation, the individual red corpuscles can be distinguished in the veins, while in the arteries they cannot, as at all times they shoot past the eye too swiftly. The fundamental observation now is verified that the blood is incessantly moving out of the arteries, through the capillaries, into the veins.

Behavior of the Red Corpuscles.-Capillaries will readily be found in which the red corpuscles move two or three abreast, or only in single file. They generally go with their long diameters parallel to, or moderately oblique to, the current. In no case will any blockade of corpuscles occur, so long as the parts are normal. The numerous red corpuscles are seen to be well fitted by their softness and elasticity, as well as by their form and size, for moving through the narrow channels. They bend easily upon themselves as they turn sharp corners, but instantly regain their form when free to do so (see Fig. 11). A very common occurrence is for a corpuscle to catch upon the edge which parts two capillaries at a bifurcation of the network. For some time the corpuscle may remain doubled over the projection like a sack thrown across a horse's back; but, after oscillating for a while, it will be disengaged, at once return to its own shape, and disappear in one of the two branches

FIG. 11.-To illustrate the behavior of red corpuscles in the capillaries: the arrows mark the course of the blood: a, a "saddle-bag" corpuscle; b, a corpuscle bending upon itself as it enters a side branch.

FIG. 12. To illustrate the deformity produced in red corpuscles in passing through a capillary of a less diameter than themselves.

(see Fig. 11). It is instructive to watch red corpuscles passing in single file through a capillary the calibre of which, at the time, is actually less than the shorter diameter of the corpuscles. Through such a capillary each corpuscle is squeezed, with lengthening and narrowing of its soft mass, but on emerging into a larger vessel its elasticity at once corrects even this deformity; it regains its form, and passes on (Fig. 12).

Evidences of Friction.-In the minute vessels, capillary and other, certain appearances should carefully be observed which are the direct ocular evidence of that friction which we shall find to be one of the prime forces concerned in the blood-movement, to which it constitutes a strong resistance. If, in a channel which admits three red corpuscles beside one another, three be observed when just abreast, it will be found that very soon the middle one forges ahead, indicating that the stream is swiftest at its core. This is because the friction within the vessel is least in the middle, and progressively greater outward to the wall (Fig. 13). In the small veins the signs of friction are

VOL. I.-6

strikingly seen, as the outer layers among the numerous corpuscles lag conspicuously. In the arterioles similar phenomena are seen if the normal swiftness of movement become sufficiently retarded for the individual corpuscles to be visible.

FIG. 13.-To illustrate the forging ahead of a corpuscle at the centre of the blood-stream. The arrow marks the direction of the blood.

FIG. 14.-The inert layer of plasma in the small vessels.

An appearance which also tells of friction is that of the so-called "inert layer" of plasma.' In vessels, of whatever kind, which are wide enough for several corpuscles to pass abreast, it is seen that all the red corpuscles are always separated from the profile of their channel by a narrow clear and colorless interval-occupied, of course, by plasma. This is caused by the excess of the friction in the layers nearest to the vascular wall (see Fig. 14). The friction thus indicated, other things being equal, is less in a dilated than in a contracted tube; and less in a sluggish than in a rapid stream. It probably varies also with changes of an unknown kind in the condition of the cells of the vascular wall.

Behavior of the Leucocytes.-If the behavior of the leucocytes be watched, it will be seen to differ markedly from that of the red corpuscles, at least when the blood-stream is somewhat retarded, as it so commonly is under the microscope. Whereas the friction within the vessels causes the throng of red corpuscles to occupy the core of the stream, the scantier leucocytes may move mainly in contact with the wall, and thus be present freely in the inert layer of plasma. Naturally their progression is then much slower and more irregular than that of the red disks. Indeed, the leucocytes often adhere to the wall for a while, in spite of shocks from the red cells which pass them. Moreover, the spheroidal leucocyte rolls over and over as it moves along the wall in a way very different from the progression of the red disk, which only occasionally may revolve about one of its diameters. A leucocyte entangled among the red cells near the middle of the stream is seen generally not only to move onward but also to move outward toward the wall, and, before long, to join the other leucocytes which are bathed by the inert layer of plasma. It is due solely to the lighter specific gravity of the leucocytes that, under the forces at work within the smaller vessels, they go to the wall, while the denser disks go to the core of the current. This has been proved experimentally by driving through artificial capillaries a fluid having in suspension particles of two kinds. Those of the lighter kind go to the wall, of the heavier 1 Poiseuille: "Recherches sur les causes du mouvement du sang dans les vaisseaux capillaires," Académie des Sciences-Savans étrangers, 1835.

kind to the core, even when the nature and form of the particles employed are varied.1

Emigration of Leucocytes.-It has been said that a leucocyte may often adhere for a time to the wall of the capillary, or of the arteriole or venule, in which it is. Sometimes the leucocyte not only adheres to the wall, but passes through it into the tissue without by a process which has received the name of "emigration." A minute projection from the protoplasm of the leucocyte is thrust into the wall, usually where this consists of the soft cementsubstance between the endothelial cells. The delicate pseudopod is seen presently to have pierced the wall, to have grown at the expense of the main body of the cell, and to have become knobbed at the free end which is in the tissue. Later, the flowing of the protoplasm will have caused the leucocyte to assume something of a dumb-bell form, with one end within the blood-vessel and the other without. Then, by converse changes, the flowing protoplasm comes to lie mainly within the lymph-space, with a small knob only within the vessel; and, lastly, this knob too flows out; what had been the neck of the dumb-bell shrinks and is withdrawn into the cell-body, and the leucocyte now lies wholly without the blood-vessel, while the minute breach in the soft wall has closed behind the retiring pseudopod. This phenomenon has been seen in capillaries, venules, and arterioles, but mainly in the two former. It seems to be due to the amoeboid properties of the leucocytes as well as to purely physical causes. Emigration, although it may probably occur in normal vessels, is strikingly seen in inflammation, in which there seems to be an increased adhesiveness between the vascular wall and the various corpuscles of the blood.

Speed of the Blood in the Minute Vessels.-As a measure of the speed of the blood in a vessel, we may fairly take the speed of the red corpuscles. It must, however, be remembered that as the friction increases toward the wall, the speed of the red corpuscles is least in the outer layers of blood, and increases rapidly toward the long axis of the tube. At the core of the stream the speed may be twice as great as near the wall. As we have seen, the stream of red corpuscles in an arteriole is rapid and pulsating. In the corresponding venule, which is commonly a wider vessel, the stream is less swift, and its pulse has disappeared. In the capillary network between the two vessels the speed of the red corpuscles is evidently slower than in either arteriole or venule; and here, as in the veins, no pulse is to be seen; the pulse comes to an end with the artery which exhibits it. In one capillary of the network under observation the movement may be more active than in another; and even in a given capillary irregular variations of speed at different moments may be observed. Where two capillaries in which the pressure is nearly the same are connected by a cross-branch, the red corpuscles in this last may sometimes even be seen to

1A. Schklarewsky: "Ueber das Blut und die Suspensionsflüssigkeiten," Pflüger's Archiv für die gesammte Physiologie, 1868, Bd. i. S. 603.

For the literature of emigration see R. Thoma: Text-book of General Pathology and Pathological Anatomy, translated by A. Bruce, 1896, vol. i. p. 344.

oscillate, come to a standstill, and then reverse the direction of their movement, and return to the capillary whence they had started. Naturally, no such reversal will ever be seen in a capillary which springs directly from an artery or which directly joins a vein. It will be remembered, however, that any apparent speed of a corpuscle is much magnified by the microscope, and that therefore the variations referred to are comparatively unimportant. We may, in fact, without material error, treat the speed of the blood in the capillaries which intervene between the arteries and veins of a region as approximately uniform for an ordinary period of observation, as the minute variations will tend to compensate for one another. This speed is sluggish, as already noted. In the capillaries of the web of the frog's foot it has been found to be about 0.5 millimeter per second. The causes of this sluggishness will be set forth later. That the very short distance between artery and vein is traversed slowly, deserves to be insisted on, as thus time is afforded for the uses of the blood to be fulfilled.

Capillary Blood-pressure. The pressure of the blood against the capillary wall is low, though higher than that of the lymph without. This pressure is subject to changes, and is readily yielded to by the elastic and delicate wall. From these changes of pressure changes of calibre result. The microscope tells us less about the capillary blood-pressure than about the other phenomena of the flow; but the microscope may sometimes show one striking fact. In a capillary district under observation, a capillary not noted before may suddenly start into view as if newly formed under the eye. This is because its calibre has been too small for red corpuscles and leucocytes to enter, until some slight increase of pressure has dilated the transparent tube, hitherto filled with transparent plasma only. This dilatation has admitted corpuscles, and has caused the vessel to appear.

That the capillary pressure is low is shown, moreover, by the fact that when one's finger is pricked or slightly cut, the blood simply drips away; that it does not spring in a jet, as when an artery of any size has been divided. That the capillary pressure is low may also be shown, and more accurately, by the careful scientific application of a familiar fact: If one press with a blunt lead-pencil upon the skin between the base of a finger-nail and the neighboring joint, the ruddy surface becomes pale, because the blood is expelled from the capillaries and they are flattened. If delicate weights be used, instead of the pencil, the force can be measured which just suffices to whiten the surface somewhat, that is, to counterbalance the pressure of the distending blood, which pressure thus can be measured approximately. It has been found to be very much lower than the pressure in the large arteries, considerably higher than that in the large veins, and thus intermediate between the two; whereas the blood-speed in the capillaries is less than the speed in either the arteries or the veins. The pressure in the capillaries, measured by the method just described, has been found to be equal to that required to sustain against gravity a column of mercury from 24 to 54 milli