PRESIDENT S ADDRESS — SECTION L. 275 



A second equally important point is that this combination is 

 unstable. Calmette (1907) has shown that whilst snake venom 

 withstands a temperature of 80" C, antivenin is destroyed at 

 68-70", especially in presence of a dilute acid. In a similar man- 

 ner the venin-antivenin complex, if submitted to the joint action 

 of acid and alcohol, is broken up, the to^xin goes into solution 

 m the alcohol, whilst the antitoxin is inactivated. 



. Danysz has also shown that pancreatic juice will restore the 

 toxicity to an innocuous mixture of ricin and antiricin by acting, 

 more quickly on the antitoxin than on the toxin. Again, it" to 

 some red blood corpuscles we add an inactivated hsemolytic serum, 

 they will become sensitized, and now only need the addition of a 

 little fresh serum, i.e., complement, to become hsemolysed. But if 

 before adding complement, we add to the sensitized cells some 

 antl-hasmolysin, i.e., serum prepared by immunizing an animal 

 to a haemolysin, then the addition of ccmplement has no effect, for 

 the hsemolysm has been neutralized by the anti-hgemol)sin. 



Of what nature, then, is the comlTination 'I Ehrlich himself 

 likened it to the union of a strong acid and a strong base accord- 

 ing to the law of definite proportions. But an observation of 

 Ehrlich's own — known as Ehrlich's phenomenon — pointed out 

 great difficulties in the way of believing this. Suppose it is found 

 that .01 c.c. of diphtheria toxin will kill a guinea pig in a given 

 time, then Ic.c. of the toxin will contain 100 units of toxin, and 

 if the concentration of antitoxin is so arranged that 1 c.c. of anti- 

 toxin exactly neutralizes 1 c.c. of toxin, then 1 c.c. of antitoxin 

 also contains 100 units of antitoxin. Then a mixture of 1.01 c.c. 

 (or 101 units) of toxin and 1.00 c.c. (or 100 units) of antitoxin 

 ought to be fatal for a guinea pig in the unit time. Strangei to 

 relate, it is harmless, and so is any mixture until there is an excess 

 of toxin over antitoxin of 25 to 35 units, when the mixture 

 (1.25 c.c. toxin and 1 c.c. antitoxin) again is fatal to a guinea 

 pig. To meet this surprising result Ehrlich suggested that the 

 toxin was really made up of two bodies, toxin and toxoii, the 

 former having a very strong affinity for antitoxin and being at 

 the same time toxic, the latter having a weak affinity for anti- 

 toxin and being non-toxic. Thus, if 100 units of '" toxin " con- 

 tains really 75 units of true toxin and 25 units of toxon, 75 units 

 of antitoxin would be able to neutralize this amount by uniting 

 with the 75 units of true toxin, leaving the 25 units of toxon 

 free but devoid of toxicity. Here again, however, the hypothesis 

 is unsatisfactory, for if 75 units of antitoxin neutralize 100 units 

 " toxin " (of which 75 are true toxin and 25 are toxon left free), 

 then 50 units antitoxin added to' 100 units " toxin " should unite 

 with 50 units true toxin, leaving 25 units toxon free and also 25 

 units toxin fres, and this last amount should be fatal for twenty- 

 five guinea pigs; hiif this is not so. As Bordet has shown, a much 

 1084.— 18 



