QUANTITATIVE ANALYSIS. 135 



weighed, is mixed with twice its volume of a saturated solu- 

 tion of sulphate of soda, and thrown upon a filter which has 

 been carefully weighed and moistened with distilled water, 

 and .also, just before receiving the mixture of blood and 

 sulphate of soda, with the saline solution. The fluid which 

 passes through should be about the color of the serum ; if a 

 few corpusles pass at first, the liquid should be poured back 

 until it becomes clear. The funnel is then covered, and the 

 fluid allowed to separate, the blood-corpuscles being retained 

 on the filter. The filter and funnel are then plunged several 

 times into a vessel of boiling water, by which all the sulphate 

 of soda which remains is washed out, and the corpuscles are 

 coagulated without changing in weight. The funnel should 

 be again covered and the water allowed to drip from the 

 filter, after which it is weighed, deducting the weight of the 

 moist filter previously obtained, which gives us the weight 

 of the corpuscles. We obtain the proportion of corpuscles 

 to 1,000 parts of blood by the following formula : 



Defibrinated blood used : Corpuscles : : Defibrinated 

 blood per 1,000 : Corpuscles per 1,000. 



The next step is to estimate the quantity of albumen in 

 the serum, and thence its proportion in the blood. For this 

 purpose we first ascertain the quantity of serum in 1,000 

 parts of blood, which is done by subtracting the sum of the 

 fibrin and corpuscles per 1,000 from 1,000. Having done this, 

 and waited ten or twelve hours for specimen !Nb. 2 to sepa- 

 rate completely into clot and serum, we take a small quan- 

 tity of the serum, about half an ounce, weigh it carefully, 

 and add suddenly twice its volume of absolute alcohol. The 

 albumen will be thrown down in a grumous mass, and the 

 whole is thrown upon a filter, which has been previously 

 moistened with alcohol and weighed. The funnel is imme- 

 diately covered, and the fluid separates from the albumen 

 very rapidly. We ascertain that no fluid albumen passes 

 through the filter by testing the fluid with nitric acid. After 



