222 BLOOD. 



first rules of analytical chemistry is, that liquid and volatile fluids 

 which are to be submitted to quantitative analysis should never 

 be allowed to stand, if it can be avoided, even to be weighed in 

 open vessels ; hence a specimen of blood intended for quantitative 

 analysis should never be allowed to coagulate in an open vessel, 

 and then perhaps to stand for 24 hours. But in following this 

 analytical rule, direct experiments show us that we obtain less 

 fibrin from the whipped blood than from washing the clot. We 

 have already shown, in the first volume, that even the fibrin 

 obtained by whipping, since it can be only imperfectly washed, 

 is never pure fibrin ; and this is far more the case with fibrin ob- 

 tained from the clot ; while a little blood-pigment always remains 

 in the former, the latter contains colourless corpuscles and the 

 walls as well as the granular contents of the coloured cells. The 

 colourless cells and the walls of the coloured cells may often occur 

 in such quantities as entirely to falsify the number representing the 

 fibrin : indeed, in the blood of the hepatic veins we have already 

 become acquainted w r ith a case in which scarcely any fibrin occurs, 

 and where it was merely the cell-membranes of the blood- 

 corpuscles that were mistaken for fibrin. The greater number of 

 the fine flakes which in whipped blood penetrate the linen filter, 

 are cell-membranes of this sort, the flakes of fibrin being in fact 

 comparatively few ; the pseudofibrin of the blood of the hepatic 

 veins passes almost entirely through the linen filter. Hence, in an 

 analysis, we have the two alternatives, either of losing some of 

 the fibrin or of simultaneously including in the calculation both 

 colourless blood-corpuscles and cell-walls ; hence more fibrin will 

 invariably be obtained from coagulated blood, in which these 

 elements are firmly enclosed by the fibrin, than from whipped 

 blood, whose fibrin (if no water has been added) had been separated 

 by a linen filter. Moreover, linen filters are not to be trusted for 

 a quantitative analysis ; for they either allow a number of minute 

 flakes of fibrin to pass through them (whether the fibrin be 

 obtained by whipping, or by kneading and pressing the clot), or 

 they become invested with a fine viscid crust of cell- walls, by 

 which even the widest meshes of the linen become clogged up. 

 Hence, in making as accurate an analysis of the blood as possible, 

 a linen filter should be altogether avoided for the quantitative 

 determination of any of the constituents, and we then find that 

 the fibrin obtained from the clot does not exceed that which is 

 obtained by whipping the blood; and further, that the experi- 

 ments instituted by Marechal, and more recently repeated by 



