294 THE CIRCULATION IN THE BLOOD-VESSELS [oil. XXI. 



move along, mostly in single file, and bend in various ways to 

 accommodate themselves to the tortuous course of the capillary, but 

 instantly recover their normal outline on reaching a wider vessel. 



At the circumference of the stream in the larger capillaries, and 

 in the small arteries and veins, there is a layer of blood-plasma in 

 contact with the walls of the vessel, and adhering to them, which 

 moves more slowly than the blood in the centre. Anyone who has 

 rowed on a river will know that the swiftest current is in the 

 middle of the stream. The red corpuscles occupy the middle of the 

 stream and move with comparative rapidity ; the colourless corpuscles 

 run much more slowly by the walls of the vessel; while next to the 

 wall there is a transparent space in which the fluid is at comparative 

 rest (the so-called " still layer ") ; if any of the corpuscles happen to 

 be forced within it, they move more slowly than before, rolling lazily 

 along the side of the vessel, and often adhering to its wall. Some- 

 times, when the motion of the blood is not strong, many of the white 

 corpuscles collect in a capillary vessel, and for a time entirely prevent 

 the passage of the red corpuscles. 



When the peripheral resistance is greatly diminished by the 

 dilatation of the small arteries, so much blood passes on from the 

 arteries into the capillaries at each stroke of the heart, that there is 

 not sufficient remaining in the arteries to distend them. Thus, the 

 intermittent current of the ventricular systole is not converted into 

 a continuous stream by the elasticity of the arteries before the capil- 

 laries are reached ; and so intermittency of the flow occurs both in 

 capillaries and veins, and a pulse is produced there. The same pheno- 

 menon may occur when the arteries become rigid from disease, and 

 when the beat of the heart is so slow or so feeble that the blood at 

 each cardiac systole has time to pass on to the capillaries before the 

 next stroke occurs ; the amount of blood sent out at each stroke is 

 then insufficient to properly distend the elastic arteries. 



It was formerly supposed that the occurrence of any transudation 

 from the interior of the capillaries into the midst of the surrounding 

 tissues was confined, in the absence of injury, strictly to the fluid 

 part of the blood; in other words, that the corpuscles could not 

 escape from the circulating stream, unless the wall of the containing 

 blood-vessel was ruptured. Augustus Waller affirmed, in 1846, that 

 he had seen blood-corpuscles, both red and white, pass bodily through 

 the wall of the capillary vessel in which they were contained ; and 

 that, as no opening could be seen before their escape, so none could 

 be observed afterwards so rapidly was the part healed. But these 

 observations did not attract much notice until the phenomenon was 

 rediscovered by Cohnheim in 1867. 



Cohnheim's experiment was performed in the following manner : 

 A frog is curarised; and the abdomen having been opened, a portion 



