1909.] On the Absorption of Agglutinin, hy Bacteria, etc. 



193 



B and C have been plotted out, it is clearly seen tliat it is quite impossible to 

 draw a straiglit line through more than two consecutive points, but that the 

 line drawn tlirough all the points is actually a curve deviating more and more 

 from the straight line, and away from the abscissa as the concentration 

 of the agglutinating serum is increasing. This is caused by the fact that from 

 the very outset C does not increase as rapidly as it should according to 

 Arrhenius' formula. Also it is seen from this and the other experiments 

 how impossible it is to find a constant exponent n if A; has to be a 

 constant, or vice versd, as demanded by Arrhenius' formula C = IcB". 

 This is well shown in the following table of figures calculated from 

 Experiments 7 and 14 (Table I). 



On the other liand, it is possible in experiments which only cover a short 

 range, to draw a line which will allow the determination of an n and k 

 which will give calculated values for B approximating to those observed — but 

 with a regular periodicity in the deviation. 



A similar periodicity in the change of value of k is also to be traced in the 

 experiments of Eisenljerg and Volk, and of Arrhenius and Morgenroth, 

 though with some difficulty owing to the great distance between the points 

 ■experimentally determined. 



In the different tables (II to XV) we have calculated from point to point 

 the exponent n, and it exhibits enormous variations in size according to the 

 region examined in each experiment. 



In Experiment 14 (Table XV), for example, we find that n will have a 

 value of about 0 8 in the weakest concentration of the serum. From this 

 point it will gradually decrease in size to zero, and then, becoming negatiA^e i 

 in sign, increase towards infinity. That this behaviour does not depend on the 

 strain of bacillus or the brand of agglutinating serum is proved by any or all 

 of the experiments undertaken with various bacteria and their homologous 

 agglutinin (see Experiments 1, 2, and 5 to 14, Tables II, III, and VI to XV). 



Experiment 5 (Table VI) proves that the variation in n is not caused by 

 the presence of the quantities of formalin added to the emulsions. 



To make it clear that this deviation could not be explained by the presence 

 of " agglutinoids " in the serum, or substances analogous to such bodies in 

 the bacteria, experiments which gave absolutely similar results were under- 

 taken with heated serum and unheated Coli bacilli filtrate (" toxin ") 

 (Experiments 3 and 4, Tables IV and V), and with heated bacilli and 

 unheated serum (Experiments 1 and 6, Tables II and VII). Now, if any such 

 bodies had been formed by the action of heat, the deviation from a straight 

 line of the curve obtained by plotting out log. B and log. C should become 

 more marked. That this, however, is not the case, is readily seen. 



