THE COMPLEMENT 221 



ferent constituents of eorpuselcs, usiii<^ — (1) pure crystalline hemo- 

 globin; (2) proteins of the stroma soluble in salt solutions; (3) an 

 extract with alcohol-ether; and (4) an extract in 1.5 per cent, sodium 

 bicarbonate. Only the last gave positive results, and the serum was 

 almost devoid of agglutinative ])roperties. Injection with corpuscles 

 that liad been digested witli trypsin gave about the same results as 

 alkaline extracts; corpuscles digested by pepsin gave a much weaker 

 serum ; in neither was agglutination obtained. According to Bang 

 and Forssmann '- ethereal extracts of red corpuscles give rise to pro- 

 duction of hemolysins on imnumization, and this "lysinogen" sub- 

 stance can be precipitated with acetone, is insoluble in alcohol, is not 

 destroyed by boiling, and gives rise to no agglutinin. Ford and Hal- 

 sey *^ obtained serum with both lytic and agglutinative powers by in- 

 jecting either the stroma or the laked blood free of stroma ; results 

 with pure hemoglobin were indefinite. Stewart *^ obtained similar re- 

 sults by immunizing with corpuscles laked by physical means, by 

 serums, or by saponin. According to Guerrini,*^ nucleoprotein ob- 

 tained from dog's blood will give rise to specific hemolysins, and Beebe 

 states that nucleoproteins from visceral organs do not have this effect. 

 Levene's alkaline extracts probably also contained nucleoproteins. 

 Immunization with extracts of tissues and cells of various sorts, even 

 when entirely free from blood (e. g., spermatozoa), may produce 

 hemolytic sera. The fact that various tissues from many different 

 species of animals, when used as antigen, may give rise to hemolysin 

 for sheep corpuscles, is an interesting but so far unexplained phe- 

 nomenon, which is discussed under "Specificity" in the preceding 

 chapter. 



The Complement. — Hemolj'tic complement possesses the same 

 properties as bacteriolytic complement, resembling enzymes to the ex- 

 tent that it is susceptible to heat and causes a disintegration of cells, 

 and is largely retained by Berkefeld filters.*" The joint action of 

 amboceptor and complement is strikingly like the activation of tryp- 

 sinogen by kinase. On the other hand, hemolysis by serum is quite 

 different from the effect of trypsin on corpuscles, as trypsin completely 

 disorganizes the hemoglobin and destroys the stroma, while in hemo- 

 lysis the stroma and hemoglobin seem to be merely separated from one 

 another but not chemieallj' altered. Again, hemolysin acts quanti- 

 tatively, although that may be due to a difference in the way the bind- 

 ing to the cell occurs, rather than in the method of action of the com- 

 plement. Landsteiner and others have suggested that a lipoidal 

 complement dissolves the corpuscle lipoids, liberating the hemoglobin, 

 while Neuberg and others haA^e supported the hypothesis that comple- 



42 Hofmeister's Beitr., 1906 (8), 238. 



— 43 Jour. Med. Research. 1004 (11), 40.3. 



- 44Amer. Jour, of Phvsiol., 1004 (11). 250. 

 45Puv. crit. di din. liied.. 1003 (4), 561. 



4GMuir and Browning, Jour. Path, and Bact., 1009 (13), 232. 



