RELATION OF TOXINES TO ANTITOXINES. 55 



14'55 c.c. of the arbitrary toxic unit, or this unit corresponds to 0'069 c.c. 

 of antitoxin. 



The values of G and x calculated from this equation are in very close 

 agreement with the observed results. 



When large quantities of antitoxine are used, a state of equi- 

 librium between the toxine and antitoxine is only reached very 

 slowly. At the same time the toxine becomes weaker, so that 

 x becomes greater. 



An analogous result was obtained in an experiment in which 

 the toxine was replaced by ammonia, and the antitoxine by boric 

 acid. Here, too, the neutralising effect of the boric acid was 

 exerted to such an extent that the law held good : 



(Free Ammonia) (Free Boric Acid) = K (Combined Boric Acid) 2 . 



It follows, then, from these experiments that the same laws 

 apply to tetanolysine and its saturation by its antitoxine as 

 govern the saturation of bases by means of weak acids. Parts of 

 the free components are invariably left uncombined. 



This was now made the starting point for attacks on Ehrlich's 

 theory of "poison-spectra." 



If we plot the curve of the neutralisation of ammonia by 

 means of boric acid, we find that the first amount added neutral- 

 ises 50 per cent, of ammonia, the next only an additional 16-7 

 per cent., and the third 8-3 per cent, more, and from this we 

 could draw the conclusion that the first aliquot portion of the 

 neutralising solution meets with a much more toxic ammonia 

 than the later portions i.e., that ammonia consists of different 

 toxic parts, which enter into combination with different degrees 

 of affinity, these standing in simple relationship towards each 

 other. This is obviously opposed to the views of EHRLICH on 

 proto-, deutero-toxines, &c. 



Hence, according to ARRHENIUS and MADSEN, we must not 

 employ as analogies for tetanolysine the conditions that govern 

 the neutralisation of strong bases and acids, as EHRLICH has 

 done in the case of diphtheria virus. In the case of strong acids 

 and bases the amount oifree components is extraordinarily small, 

 and, practically, we have only to take into account the compound 

 and the excess of one of the components. Measured by this 

 standard, then, the neutralisation curve of diphtheria virus 

 deviates, as EHRLICH has shown, very considerably from the 

 simple curve of the neutralisation curve of acids and bases, and 

 lends support to the view of the existence of different poisonous 

 constituents. But if we take as the standard the conditions of 

 equilibrium that occur in the case of weak affinity, conditions 



