VIII 



with those occurring when a body is thrown upwards 

 and then owing to the gravitation, again falls to the 

 ground. 



The difference between an actively and passively 

 immunised organism is, that in the latter case, the anti- 

 body will decrease regularly to 0, while in the former 

 case the antibody after a certain time usually reaches 

 an equilibrium or continues to decrease very slowly. 

 At this point it seems to be most probable that there 

 is a production of antibody, which covers the loss. It 

 appears trom the antitoxin and agglutinin curves, that 

 an actively immunised organism, in every case for a 

 certain time, is liable to keep such an equilibrium, 

 which it tries to restore if disturbed, as for instance 

 by a copious bleeding. That this production is not re- 

 cognisable during the third phase, might probably de- 

 pend on its relatively small magnitude. 



If the immunisation is produced by daily injections of 

 small amounts of culture, quite a similar curve is obser- 

 ved (see f. inst. curve No. 3). Here the remarkable fact 

 may be noticed, that the fall occurs in spite of the con- 

 tinuing injection of culture. The exact concordance 

 between the observed and calculated values of aggluti- 

 nin, which is perceived during this fall from its very 

 beginning, indicates that the production has quite ceased, 

 the cells probably being fatigued and no longer capable 

 to react on the injections. 



In the case of the goat, curve 3a, a negative constant 

 Cj= -130 is observed. This indicates that ourstartingpoint 

 for the time (t = 0) is wrong. Infacta glance at the curve 

 reveals that no observation was made between the w / l 

 and 21 / 1 . The acme would have been observed on the 

 2 %, and this is the real starting point from which the 

 mite t has to be counted, giving a positive value for q. 



12 



