1911.] Serum Reactions in Pulmonary Tuberculosis. 385 



1/60) might easily be realised in antigen in a short space of time which, 

 under many methods of control for fixation purposes alone, would escape 

 detection ; indeed Serum 860 was in the very earliest part of the work 

 recorded as a positive example of fixation. Since then, because of the 

 repeatedly similar results, and in the light of these conceptions and the 

 clinical correlation, the value of this is felt to be doubtful. 



(8) Natural Sheep Corpuscle Amboceptor. — This haemolysin is frequently 

 found in the serum of man and animals, and may interfere with a correct 

 interpretation of the end results. It can be avoided by first treating the 

 serum with corpuscle solution, which, however 'means an increase of technique 

 that is particularly objectionable with patients' serum for purposes of clinical 

 correlation. As a routine I have instead used two controls, so that the extent 

 of the influence of this amboceptor under the circumstances of the test may 

 be fairly accurately gauged. These are serum with in one tube the maximum 

 strength of antigen, and in the other saline ; saline solution being used to 

 replace the immune haemolysiu with the second incubation. Thus the control 

 with saline gives an end result indicative of the activity of the natural 

 haemolysin under slightly more favourable conditions than exist in the 

 protocol proper, i.e. greater saline dilution. With no haemolysis in this latter 

 tube it has been shown that slight sensitisation of corpuscles can be produced 

 by adsorption experiments. In only a few instances has the control with 

 antigen shown any haemolysis. However, no haemolysis in the antigen 

 control tube for natural amboceptor is not as satisfactory a control as might 

 at first appear, because of the underlying principles, which can be shown by 

 the apparent degree of the inhibitive reaction before, and the real reaction 

 after, adsorption of the amboceptor. This is chiefly due to the fact that the 

 curve of the end results with regard to time and degree is not in direct pro- 

 portion to increase of the different biological constituents of the test. In this 

 regard, the use made of the term one and two units is inaccurate ; such 

 nomenclature is, however, convenient, and is in this way qualified. 



In earlier parts of the work a negative result in the antigen control tube 

 was relied upon too extensively, and certain of the interpretations made were 

 probably not correct. With any more than " some " haemolysis in the saline 

 tube, adsorption of the natural amboceptor should be performed for an 

 accurate estimation of the amount of inhibitin. 



The presence of natural amboceptor seems to explain certain irregularities 

 of the inhibitive reaction obtained with varying amounts of serum and anti- 

 complementary antigen. Thus a serum may show no evidence of inhibitin 

 with 0T cc. and a full reaction with - 3 c.c, while the controls show the 

 presence of the natural haemolysin. After adsorption, the serum would show 



