23 



this changing of colour may produce on the co- 

 lourless fluid, which is conveyed by the capillary 

 vessels. As the colouring- matter is not dissolved, 

 but only mechanically mixed with the blood, we 

 may consider the separation of the arterial blood 

 into coloured and uncoloured, as a process of fil- 

 tration, which only admits the colouring matter 

 to penetrate into the veins, while, on the other 

 hand, the serum alone is forced into the finest 

 channels, 



It would be of the greatest importance to Ani- 

 mal Chemistry, to examine the serum, in the 

 state in which it penetrates the capillary vessels ; 

 but I cannot conceive how it would be possible 

 to collect it, even in a very small quantity. It 

 is probable, that this humour is the common se- 

 rum, which also contains dissolved fibrin; and if 

 this supposition be correct, it follows, that the 

 fibrin within the vessels must also be dissolved in 

 the serum, and not belong to the suspended co- 

 louring matter. Such I have always considered 

 the composition of the blood ; but, except the 

 confirmation we receive, by examining the humour 

 of the absorbing vessels, I have not been able to 

 discover any experiment to confirm, or refute, 

 this supposition. 



