470 IMMUNITY 



immunised against ricin by feeding as described above, could 

 protect another mouse against forty times the fatal dose of that 

 substance. He considered that in the case of the two poisons, 

 antagonistic substances " anti-ricin " and " anti-abrin " were 

 developed in the blood of the highly-immunised animals. A 

 corresponding antagonistic body, to which Fraser has given the 

 name "antivenin," appears in the blood of animals in the process 

 of immunisation against snake poison. 



Theso investigations are specially instructive, as such vegetable 

 and animal poisons, both as regards their local action and the 

 general toxic phenomena produced by them, present, as we have 

 seen, an analogy to various toxins of bacteria. 



Nature of Antitoxic Action. This subject is only part of the 

 general question with regard to the relation of anti-substances 

 to their corresponding substances, but it is with regard to anti- 

 toxic action that most of the work has been done. We have to 

 consider here two points, viz. (a) the relation of antitoxin to 

 toxin, and (b) the source of the antitoxin. With regard to the 

 former subject there has been much diversity of opinion, but 

 the evidence now available goes to show that the antagonism 

 between toxin and antitoxin is not a physiological one but that 

 the two bodies unite in vitro to form a compound inert towards 

 the living tissues, there being in the toxin molecule an atom 

 group which has a specific affinity for the antitoxin molecule or 

 part of it. We shall consider the facts in favour of this view, 

 and in doing so we must also take into account the anti-sera of 

 the vegetable toxins, of snake poisons, etc. 



When toxin and antitoxin are brought together in vitro it 

 can be proved that their behaviour towards each other resembles 

 what is observed in chemical union. Thus it has been found 

 that a definite period of time elapses before the neutralisation 

 of the toxin is complete, that neutralisation takes place more 

 rapidly in strong solutions than in weak, and that it is hastened 

 by warmth and delayed by cold. C. J. Martin and Cherry, and 

 also Brodie, have shown, that in the case of diphtheria toxin and 

 in that of 'an Australian snake poison the toxin molecules will 

 pass through a colloid membrane (p. 166), whilst those of the 

 corresponding antitoxin will not. Now if a mixture of 

 equivalent parts of toxin and antitoxin is freshly prepared and 

 at once filtered, a certain amount of toxin will pass through, but 

 the longer such a mixture is allowed to stand before nitration 

 the less toxin passes, till a time is reached when no toxin is 

 found in the filtrate. Further, if the portion of fluid which at 

 this stage has not passed through the filter be injected into an 



