1909. J On the Absorption of Agglutinin by Bacteria, etc. 195 



absorbed, but on the varying amounts of both, the presence of albumen, 

 dilution of serum, presence of salts, temperature, time of reaction, and other 

 variable factors. 



4. Turning to the fourth point brought forward by Arrhenius, that the 

 agreement is satisfactory between observed figures and those calculated 

 according to his formula C = /vB" and published in his papers, we consider 

 that the accordance between his figures is by no means good. Arrhenius 

 himself regards the deviations he finds as entirely due to experimental error, 

 because he was informed by Eisenberg and Volk that such great variations 

 were quite possible in the technique that they had used. To us it is clear 

 that, leaving alone the great experimental error, all the experiments 

 calculated by Arrhenius show a certain periodicity in the increase and 

 decrease of k in the same direction as found in our own experiments, but 

 not taken into consideration by him. Our own results show that, even if 

 we chose the best possible n, B calculated and B observed will only agree 

 approximately within a certain small range. Beyond these limits the 

 values for B observed will be much bigger than for B calculated, a natural 

 result of what has been stated previously, that C does not increase so fast as 

 it should if Arrhenius' formula C = A;B" were correct. 



As we have proved beyond dispute the great and regular variations 

 occurring in the " constants " n and k of Arrhenius' formula C = /cB" when 

 applied to the absorption of agglutinin by bacteria, we record no figures 

 calculated according to that expression, since it is now obvious how 

 absolute must be the disagreement between the observed values of B and 

 those thus calculated. 



At a later date it is our hope to deal with the mathematical considerations 

 arising in connection with the figures in this paper. 



