PHYSIOLOGY IN THE STUDY OF DISEASE 93 



most natural idea would be to inject blood into the 

 depleted vessels. This transfusion of blood has 

 been used for a long time to make good great losses 

 of blood by haemorrhage. Unfortunately the 

 blood of lower animals cannot be safely used, since 

 it is injurious to the blood of the receiver. It is 

 often dijQ&cult to get human blood, and the use of 

 this is also not without danger. Great care must 

 be exercised in determining the character of the 

 blood of the donor. Plain water cannot be used, 

 for it destroys the cells of the blood. But a solution 

 of salt may be made which does not do so, and this 

 has been tried. It certainly raises the blood 

 pressure, but only for a very brief period, because 

 it exudes from the vessels and the patient's condi- 

 tion is not permanently relieved. 



Bayliss reasoned that the fluid which must be 

 introduced must resemble blood in its more im- 

 portant physical characters, and must not produce 

 any prejudicial effect. Much time had been spent 

 by physicists and physiologists in the study of what 

 appeared to be matters of purely scientific interest, 

 the exact physical condition of the colloidal state 

 of matter, and of the phenomena of osmosis and of 

 viscosity. From his knowledge of these studies 

 Bayliss drew the conclusion that a solution of gum- 

 arabic of a certain strength should fulfil the conditions. 

 He tried it first upon animals, and found that its 

 introduction raised the blood pressure and main- 



