1911.] Serum Reactions in Pulmonary Tuberculosis. 389 



Apparently, reproduction of the fixative strength of serum may take 

 place hy the addition of fresh serum. 



The following results in this connection have been obtained : — 



February 16, 1911. — Fresh serum from two guinea-pigs killed on 

 February 15, 1911, standardised 1/20 dilution to 1 unit; used satisfactorily 

 in a 1/10 dilution in the succeeding protocol. 



February 18, 1911. — About 12 c.c. of fresh guinea-pig serum (drawn 

 February 17, 1911) added to between 3 and 4 c.c. of serum remaining from 

 the preceding lot. Immediately after adding the fresh to the older serum 

 the mixture standardised at barely 1/20. Within 3 hours this was used 

 in a 1/10 dilution. The protocol of 18 sera contained well-marked examples 

 of fixative positive sera, and yet nothing less than partial haemolysis 

 resulted, in a few instances, with the higher strengths of antigen and 

 serum. The following morning there remained sufficient of the mixture of 

 guinea-pig sera to show that the fixative strength was over 1/50. 



The chief points that have been dealt with may be brought together as 

 follows : — 



(1) The technique is concerned with specific biological factors, but is 

 hampered also by other specific and non-specific conditions. For a correct 

 estimation of the specific effect of the antigen-serum complex upon comple- 

 ment these must be adequately controlled. 



(2) These factors may be regarded in the light of their tendency to induce 

 or hinder haemolysis. 



A. Conditions favouring no haemolysis : — 



(a) Presence of specific tuberculous sensitisers. 



(b) Non-specific effect of antigen upon complement. 



(c) „ „ serum suspension of complement. 



B. Conditions favouring haemolysis : — 



(a) Presence of specific inhibitin. 



(b) „ natural sheep corpuscle amboceptor in the tested 



serum. 



(3) Effect of saline dilution. 



(4) Variations in the fixative strength of complement. 



(5) Experimental error incident to the performance of the technique. 



