226 Professor Svantp Arrhenms [June 0, 



tion) was added to different quantities of antitoxin, and after this 

 mixture had been kept at 37° C, during so long a time that no further 

 sensible chemical change took place, it was mixed with 3 c.c. of the 

 tetanus poison. It was found that no effect is observed if the 

 quantity of antitoxin added is less than 0'16 c.c. of the solution 

 used. With higher quantities of antitoxin, the effect is proportional 

 to the used quantity diminished by 0'16 c.c, just as might be 

 expected. The quantity of antitoxin equivalent to 1 c.c. of tetano- 

 lysin was by direct experiments found to be 0"18 c.c, which corre- 

 sponds very well with the 0* IG c.c. calculated from the Danysz-effect, 

 especially if we regard the partial decomposition of the compound 

 toxin-antitoxin. 



If we had used trichloracetic acid instead of monochloracetic 

 acid, the " Danysz-effect " would have been found to take place in an 

 interval three times as great as the interval of neutralization, whereas 

 with monochloracetic acid the two intervals are equal. By the action 

 of tetanolysin on its antitoxin, the interval of the Danysz-effect was 

 found to be more than six times as great as that of neutralization. 

 This probably depends on the relatively complicated structure of the 

 lysin-molecule. 



There are some other peculiar phenomena which are to a certain 

 extent similar to the Danysz-effect. If we take a quantity of a lysin 

 which is just sufficient to cause the total hemolysis of a given 

 quantity of red blood-corpuscles, and divide this quantity in two 

 equal parts, and add them successively with an interval of some 

 minutes at 37°, to the poison, the hemolysis will not be com- 

 plete. This depends on the property of the contents of the blood- 

 corpuscles to combine with a quantity of lysin, which exceeds the 

 quantity causing total hemolysis by many, perhaps a hundred, times. 

 The first half of the red blood-corpuscles, therefore, takes away 

 nearly all the poison from the fluid, so that there remains practically 

 no poison at all for the second half of the red blood-corpuscles, 

 which, therefore, remain nearly intact, w^hereas the first half is 

 totally hemolysed. The result, therefore, is that only a little 

 more than fifty per cent, of the total quantity of blood-corpuscles 

 is hemolysed, whereas if all these corpuscles had been mixed with 

 the lysin simultaneously the hemolysis would have been complete. 



This observation was made by Bordet ; similar effects are found 

 for the agglutination of bacilli by Joos, and they may be explained 

 in the same manner. There is a small secondary effect, as Morgen- 

 roth has shown in these cases. The compound of the poison with 

 the content of the cells is partially decomposed, and therefore a 

 small part of the poison diffuses out into the liquid and attacks 

 the cells which were intact before. But this effect is a very 

 slow one, so that it does not disturb the first result very much. 



Quite the opposite effect to that observed by Bordet may be 

 stated. Suppose we have so unich tetanolysin that it is just 



