194 



Dr. G. Dreyer and Mr. J. S. C. Douglas. [Nov. 16, 



Experiment 7 (Table VIII), in which the serum before use had been treated 

 with bacteria to remove such bodies if they were present, offers additional 

 proof on this point. 



It is very interesting to note that it is not at present possible to 

 distinguish between the action of the Coli bacilli filtrate (" toxin ") and that 

 of the bacteria themselves, on agglutinin ; this is seen from Experiments 2 

 and 3 (Tables III and IV). Therefore the supposition expressed by one of 

 us (G. D.) on an earlier occasion that, as the partition law of Arrhenius did 

 not hold good for the interaction of such " toxin " and agglutinin the same 

 would most likely be found to be the case if the interaction of bacteria and 

 agglutinin were examined, was correct. 



A further point of interest is that in spite of the partition law not 

 holding good in the case of " toxin " and agglutinin, the above formula will^ 

 nevertheless, if used for calculation, give a better agreement between 

 calculated and experimental figures in this case than in that of bacteria and 

 agglutinin. 



Comparing together the absorptions, of agglutinin from heated serum 

 and from unheated serum, and plotting out the values log. B and log. C, it 

 appears that tlie deviation from the straight line is less in the case of the 

 heated serum. The higher the temperature, within certain ranges, to which 

 the serum is heated, the more constant will n be found within corresponding 

 ranges of concentration of serum, and the nearer to a straight line will lie 

 the curve obtained. Thus in Experiment 4 (Table V), where the scrum was 

 heated to 70° C, and there was an increase in concentration from about 

 4 to 60 units, nothing better than a straight line can be drawn, giving 

 an n of about 0"73, and a, k oi about I'O, yvliile in Experiment 3 (Table IV), 

 where the serum is only heated to G0° C, 7h varies from about 1"1 to ()"3. 



From the whole series of experiments it is clear that one is not justified 

 from the size of n in forming a conclusion (as done by Arrhenius) as to 

 the nature of the interaction, because n may have any size from near 

 one to minus infinity in the same experiment, depending only on the 

 concentration of the serum. This is the more the case since we have 

 found variations in the size of n of very much the same kind in the 

 " adsorjition " of agglutinins and other bodies by charcoal, as we shall 

 describe in a latci' ])a{)cr.* 



It is further seen fi'om Experiments 1 to 14 (Tables II to XV) that no 

 great stress can be laid on the actual size of n or h, since n is not alone, 

 or even mainly, dependent upon tlie kind of absorbing matter or substance 



* A |)ti'!iiiiiiiiiry ('(iiiiiiiiuiiratidii mi this snlijcct was ffivcii at tlio .January iiiiMjling of 

 the I'athohigical Socioty of (irc^at I'.i itriiii and 1 1 eland, 1009. 



