CHEMICAL COMPOSITIOX. 27 



matter of blood, is produced by the action of alkalies upon hajmoglobin, and is 



tenned lurmntin. When obtained pui-e 



this body is insoluble in water, and also Fig. 15. 



in alcohol and ether, except in presence 



of an alkali. Like hjemoglobin, its alka- ■/■ ~''y^ 



line solutions produce a different eflfect / v j^ ^4, ^_"^l;f 



upon the si^ectniin according as they are ak / ""^ 



oxidated or deoxidated ; the absoi-ption j ' y ^^ 1 *'V-' '^ 1- 



bands are, however, entirely different ^j^ ■^ -^ \ ,=, y \ 



from those of ha3moglobin. The effect of y ^^~^ ^ V" v* ^ 



acids upon hajmatin is to separate the i ''"^ ^ ^. "^^ f^ \ , 



iron and to transfonn the substance into rr^ \ ' , /^ '*^ ''"' ^ 





hffmnto'm (acid-hajmatin), the spectinun 

 of which is characterised by the presence 



of fiiHv absoi-ption bands. A compound v- ■ "'/^ _ . 



of ha3matinwithhydi-ochloricacid(/;rt;'?«;;/, 'V "^"^i" -V 4 ^^'' ' 



Teichmann) is readily obtained from '''Jj ^ ^\' /'J t, 



haemoglobin by wamaing it with a little /—^ i — 



salt and glacial acetic acid. On cooling, 'f'^/ j, _- "- 



it crystallizes out in minute reddish-brown - — 



acicular prisms (fig. 1")), the demonstra- 

 tion of which affords a positive proof Fig. 15.— Hjijun Crystals, ^Iagxified 

 of the presence of blood-colouring matter. (from rreyer). 



Tliey may readily be obtained from di'ied 



blood without the addition of .salt, merely hy wanning it with concentrated acetic 

 acid.* 



Inorganic Constituents of tlie Eed Corpuscles. — Besides the 

 iron of the h;i3mog-lul)in, the red corpuscles contain a certain ])ropor- 

 tion of salts, chiefly of potash and lime, combined with carbonic and 

 phosphoric acids. It is, however, impossible to obtain the corpuscles 

 in quantity, sufficiently isolated for exact analysis. 



Proportion of Red Corpuscles. — The red corpuscles form by far 

 the largest part of the organic matter in the blood : their ]iroportio]i 

 may be approximately ascertained by filtering defibrinated blood mixed, 

 with solution of snlj^hate of soda ; or by weighing the dried clot, and 

 making allowance for the fibrin it contains. The latter method, how- 

 ever, will serve only to give a rough estimate, as the very uncertain 

 amount of sernm remaining in the clot and affecting its weight cannot 

 be determined. Prevost and Dumas made too large a deduction for the 

 solid matter supposed to belong to the retained serum, and this 

 reduced the estimate of the dried corpuscles too much, viz. to 129 parts 

 per 1000 of blood. Lecanu also gives it at from 120 to 130 : Becquerel 

 and Ptodier at from 131 to 152. Schmidt, from three modes of cal- 

 culation, which it is needless here to explain, arrived at the conclusion 

 that the proportion of moist red corpuscles in 1000 parts of blood is 

 from 480 to 520 ; but there are reasons for regarding this as too 

 higli an estimate. Hoppe-Sevler estimates the proportion at 32 G per 

 lUOO. 



Different obsen'ers agi'ee that, as a general rule, the ijroportion of red particles 

 is gi-eater in the blood of the male sex than in that of the female. Lecanu gives 

 the following mean result, derived from numerous analyses. exhil:)iting the pro- 

 poition of diy crassamentum and water in the blood of the two sexes. No 



* For some interestinji observations by A. Gamgee on the action of nitrites on hasmo- 

 globin, see Philosophical Transactions, 1868, p. 5S9. 



