82 On Diaj)edesis. 



becomes filled with coloured corpuscles, which have a tendency to 

 reach the external layers of the fluid, whilst the colourless are 

 adhering to the walls of the vessel. Frequently they are loosened 

 and swept onwards by the current, but many remain stationary, and 

 in a short time elevations will be observed on the outer side of the 

 wall of the vessel. These gradually increase in size at the expense 

 of the corpuscles on the interior, until at last the perfect corpuscles 

 are seen wandering away from the vessel into the tissues by means 

 of their delicate hyaline processes or pseudopodia. 



III. We will now consider the fallacies, and the various 

 explanations, of these phenomena. 



Fallacies. — One great source of error is the rupture of some 

 vessel accidentally or from operation. The blood-corpuscles spread 

 over the field of the microscope, and a false impression is conveyed 

 to the mind of a superficial observer, viz. that they have escaped 

 fiom the vessel under observation; but careful focussing will soon 

 reveal the mistake. Very often a small vessel may partly cover, or 

 be concealed by, a corpuscle. Again, the observer imagines he sees 

 a corpuscle emerging from the vessel ; but no errors of this descrip- 

 tion can ever be committed if the following simple rule be rigidly 

 observed, viz. that all corpuscles on their passage through the 

 capillary walls consist of three distinct portions — one on the 

 exterior, another on the interior, and a third or neck uniting 

 them ; in fact, a distinct constriction should be seen. 



Explanations. — Observers do not agree on the manner in 

 which the corpuscles escape from the unruptured vessels. 



Cohnheim considers that by virtue of the amoeboid motion 

 they possess, their exit is readily efiected through the false stomata 

 between the endothehal elements of the vessel. He also states that 

 the molecular arrangement of the constituents of the endothehum 

 is altered ; hence it becomes sticky or viscid, and the onward course 

 of corpuscles is retarded. Billroth believes that it is due to some 

 chemical or some molecular changes producing softening of the 

 vascular walls. 



Cohnheim stated that pressure assisted materially in the process ; 

 but if the arteria-media of a rabbit's ear be exposed, the corre- 

 sponding vein opened, and distilled water injected through the 

 artery at a lower pressure than that of the blood, it will produce 

 inflammation, although no fluid has been expressed from the vessels. 



The introduction into the circulation of a small quantity of 

 2 per cent, solution of common salt will also produce an abundant 

 migration. 



Strieker and Prussak consider the process due to an " active 

 state " of the walls of the vessels, which consist of a homogeneous 

 extensile protoplasm, and adduce as proofs that all colloid sub- 

 stances allow other colloid substances to pass through without their 



