IMMUNITY REACTIONS 201 



the lecithin on the surface of the blood corpuscles. It is very prob- 

 able, as assumed by the investigators mentioned, that lecithin in this 

 instance acts as a solvent for the lipoid membrane of the blood cor- 

 puscles. Numerous similar models in which complement was re- 

 placed by lipoid were subsequently devised. 



Reversibility. The combination of agglutinin with bacteria and red 

 blood corpuscles is partially reversible; it may be partially removed 

 at a higher temperature by washing with physiological salt solution 

 (K. LANDSTEINER, EISENBERG and VOLK). 



This is deduced from an experiment of J. Joos.* He mixed typhoid 

 bacilli bearing agglutinin with untreated typhoid bacilli, and all 

 the bacilli became agglutinated. There must have been a with- 

 drawal of agglutinin from the typhoid bacilli which had been treated. 

 In a similar way, J. MORGENROTH* 1 demonstrated that the combi- 

 nation of amboceptor and red blood corpuscle is partially reversible. 

 Reversibility within the organism, where numerous varieties of cells 

 occur, is of great practical importance. 



B. The Distribution of Immune Substances Between Dis- 

 solved Colloids and Solvent. 



The colloid-chemical theories regarding the combination of toxin 

 and antitoxin are tacitly based upon the assumption that toxin and 

 antitoxin behave like a suspension or a hydrogel; they premise that 

 surfaces occur, upon which, for instance, the toxin may concentrate 

 in accordance with the laws of adsorption. Theoretical basis for this 

 assumption is lacking. Very little is known concerning the fixation 

 of crystalloids and of hydrosols by hydrosols, when no precipitate 

 occurs. There are, at present, two methods of attacking the problem. 

 By means of ultrafiltration, H. BECHHOLD* 4 has shown that the com- 

 bination of methylene-blue with serum albumin satisfies the conditions 

 of an adsorption (see p. 25). L. MICHAELIS and P. RONA have used 

 the osmotic compensation method in order to determine the kind of 

 combination in which sugar, Ca, etc., are fixed in the blood. Both 

 methods are recent and had not previously been utilized in the solu- 

 tion of this problem.* On this account I consider it unprofitable to dis- 

 cuss at present the manner in which toxin and antitoxin are combined. 



It may be mentioned that the toxin-antitoxin combination is in- 

 completely reversible in part and in part irreversible. P. EHRLICH 

 and his pupils demonstrated by numerous biological investigations, 

 that the combination between diphtheria toxin and its antitoxin 

 rapidly became irreversible. 



The relations between precipitin and precipitable substance is some- 



