512 IMMUNOLOGY 



from 0.1 iiig. to 225 mg. into a horse. A period of three montlis 

 is employed for immunization. When the precipitin titer is suffi- 

 ciently high, the horse is bled and the serum obtained for use in 

 the test. The authors find it necessary to employ the optimum 

 proportion technique of Dean which is discussed in the chapter on 

 Precipitins. They choose the Frankfurt standard tuberculin as 

 the standard for comparison, and arbitrarily assign to it a value 

 of 100 units per cubic centimeter. "The value of any other 

 tu])erculin is inversely proportional to the smallest amount re- 

 quired to produce the same degree of precipitation as the stand- 

 ard with the same quantity of serum under the same conditions. ' ' 

 In the opinion of Dreyer and VoUum the strength of a tuberculin 

 determined by the precipitin method compares favorably with 

 tlie strength determined by skin tests on human subjects. 



Complement Fixation Method of Watson and Heath. — Stand- 

 ardization of Tuberculin by Complement Fixation : This method 

 is credited to Watson and Heath. They prepare an antiserum by 

 injecting intravenously into a horse 1 mg. of virulent, moist 

 tubercle bacilli followed in fourteen days and twenty-one days 

 by 2 mg. and 3 mg., respectively, of the same kind of material. 

 The concentration of complement fixing antibody reaches a peak 

 during the seventh week. At this time the titer is said to be 

 about 0.0066 or 150 units per c.c. when titrated against a fixed 

 dose of standard tuberculin. They state that a fixed dose of 20 

 units of antibody is the most suitable one for titration of tuber- 

 culin. In standardizing commercial tuberculins they determine 

 the minimum amount of the tuberculin which, when mixed with 

 20 units of antiserum, will completely fix the unit of complement 

 used. This amount of tulierculin is said to represent one active 

 unit. 



Objections to Ijethal Tests. — Criticism of tests based upon 

 lethal doses of tuhereulin. As regards the three methods, which it 

 will be recalled involve the death of the guinea pig, there are too 

 many unknown factors that affect the outcome of the test. Among 

 these is the extent of the tuberculosis within the guinea pig. This 

 cannot be known or controlled. In Long's opinion "individual 

 variation (in animals) is so great as to rob the test of much of its 

 value. ' ' 



