QUANTITATIVE ANALYSIS. 133 



The above process is perhaps as simple and reliable as 

 any ; but of course each chemist has some slight modifica- 

 tions. By some the globules are estimated by drying the clot 

 after coagulation and deducting the weight of the fibrin. 

 Some recommend to expose the fibrin after desiccation to in- 

 cineration, and deduct the weight of the residue of inorganic 

 matter. All of the processes, however, are materially the 

 same, and differ but little from that employed by Provost and 

 Dumas. As before remarked, the results, as regards the fatty 

 and inorganic constituents of the blood, are as accurate as 

 possible with our present means of investigation ; and the 

 comparative results, in analyses of the blood for fibrin, albu- 

 men, and corpuscles in health and disease, which have crowned 

 the labors of Andral and Gavarret, Becquerel and Rodier, 

 and a number of others, are of permanent value. But a 

 glance at the process, and the quantities given for the fibrin, 

 albumen, and corpuscles, indicate that the whole is inconsist- 

 ent with our ideas of the condition under which these sub- 

 stances exist in the body. Microscopic examination shows 

 that at least one-half the mass of the blood consists of cor- 

 puscles, while analysis gives only 135 parts per 1,000. The 

 fibrin of the blood is sufficient to entangle, as it coagulates, 

 all the corpuscles, and with them form the clot ; yet we are 

 told that its proportion is 2*5 parts per 1,000. We boil the 

 serum, the albumen changes from a fluid to a semi-solid con- 

 dition, and the whole mass is solidified ; yet the estimate of 

 its proportion is YO parts per 1,000. The fact is that these 

 estimates give us only the dry residue of the organic princi- 

 ples ; and to form an idea of their actual proportion, we should 

 estimate them, if possible, with their water of composition, 

 and united with the inorganic salts, which cannot be separated 

 from them without incineration and consequent destruction. 



With this end in view, and forwant of a better process, we 

 may employ the following mode of analysis, which is easy of 

 application, and sufficiently accurate for all practical purposes. 1 



1 See an article by the author, on The Organic Nitrogenized Principles of the 



