BLOOD. 243 



The dififerences between these analj^ses are ob\ious. The 

 sohd constituents obtained by Andral and Gavarret's method are 

 8-5 higher, in 1000 parts of blood, than by mine ; moreover, the 

 quantity of corpuscles obtained by them considerably exceeds 

 the hsematoglobulin separated by my method. If we assume 

 that the 8*5 parts of water which Andral and Gavarret^s method 

 did not succeed in removing, were retained in the clot, the cor- 

 puscles would be reduced from 10886 to 98'3 ; in which case 

 the discrepancy between the two analyses would be much less 

 striking. 



1000 parts of blood would then contain : 



According to Simon. 

 Fibrin .... 2-00 



Albumen, with extractive matters, 



and salts . . . 86-40 



Haematoglobulin . . 93-60 



According to Andral and Gavarref. 



Fibrin .... 2-00 



Solid residue of serum . 80-50 



Blood-corpuscles . . 99-50 



It must, however, be remarked, that the sum of the haematin 

 and globulin, in my analyses, can never represent the absolute 

 quantity of blood-corpuscles. As has been previously remarked, 

 the nuclei and capsules of the blood-corpuscles have been esti- 

 mated as albumen by my method, as fibrin by Berzelius, and 

 as appertaining to the corpuscles by Andral and Gavarret. 



Their absolute weight has never been accurately ascertained, • 

 but it cannot be larger, since the quantities of fibrin obtained 

 by washing the clot, and by whipping fresh blood difl'er very 

 little. Further, a portion of fat separated by my method, be- 

 longs to the blood-corpuscles, and we cannot deny the possi- 

 bility of the corpuscles containing albumen. 



My analyses, moreover, aim not merely at the determination 

 of the proportion of the fibrin, of the corpuscles, and of the 

 solid residue of the serum, but they are intended to embrace 

 the determination of the most important proximate constituents 

 of the blood ; and if the haematoglobulin, or possibly the glo- 

 bulin be regarded as constituting the principal mass of the cor- 

 puscles, I can succeed in tracing their increase or decrease 

 by means of the proportion of the haematoglobulin or globulin. 



The following objections may likewise be brought against 

 Andral and Gavarret^ s method. 



' Nasse (Das Blutin mehrfacher Beziehung, &c., Bonn, 1836, p. 109) has attempted 

 to form a quantitative analysis of the nuclei. 



