218 BLOOD. 



with hot water (Figuier) or is first washed off the filter in tepid 

 water, and then coagulated by hoiling. Practicable and accurate as 

 this proceeding appears at first sight to be in theory, it is not to 

 be depended upon in practice. Notwithstanding the precautionary 

 rules recommended by Dumas, some of the blood-corpuscles almost 

 always pass through the filter, and this is the more likely to occur, 

 the more rapidly the corpuscles adhere in dark red masses to it; 

 but even when the filtered fluid appears only little coloured, we 

 can always detect plenty of corpuscles in it wjth the microscope, 

 or at all events perceive the deposition of a red sediment; the 

 fluid often passes so slowly through the filter, that the latter 

 becomes completely blocked up by the more or less -altered blood- 

 corpuscles. This method of procedure is very often inapplicable to 

 diseased blood, either from its corpuscles passing as readily through 

 the filter after the addition of sulphate of soda as before, (Didiot 

 and Dujardin*) or because the serum is so viscid and almost 

 gelatinous, that it will not pass through the filter. In a very small 

 number of cases this difficulty may be got over by substituting a 

 solution of sugar for one of sulphate of soda (Poggialef). This is, 

 however, the most important question Is all the serum actually 

 separated in this manner ? If this were the case, this method might 

 at all events be used as a check to other methods, as for instance 

 to Scherer's, and we should thus probably be able to discover a 

 coefficient for the error (consequent on the serum retained in the 

 clot) which is unavoidable in the preceding methods ; but unfor- 

 tunately this is not the case, for the corpuscles collected on the 

 filter are by no means free from serum after two or three washings 

 with a solution of Glauber's salts, as Hofle believes ; for the fluid 

 running off the corpuscles even after six or eight washings, is not free 

 from the constituents of the serum, (if indeed so many washings 

 do not cause the disintegration of the corpuscles or the clogging 

 up of the filter ;) this is the reason why, as Gorup and Hinter- 

 berger found, this method yields more dry blood-cells than any 

 other, notwithstanding the above mentioned loss of corpuscles 

 and their constituents, (which, when the globulin of the blood- 

 cells is imperfectly coagulated, remain dissolved with the sulphate 

 of soda, especially if a little acid has not been added to the fluid 

 to be coagulated). This excess of blood-cells becomes more in- 

 telligible when we have convinced ourselves (as I have often done) 

 that clear blood-serum becomes strongly turbid by a saturated 



* Compt. rend. T. 23, p. 227. 

 t Ibid. T. 25, pp. 198201. 



