480 STUDIES IN IMMUNITY. 



sequently, if sensitized corpuscles are subsequently introduced into 

 the mixture, they undergo no hemolysis. Certain other mixtures, 

 the list of which will be found in our previous articles, give the 

 necessary control; in one of them in particular it is proved that if 

 normal serum, which is not sensitizing, is used in the first mixture, 

 instead of the specific serum, there is no delay in the hemolysis of 

 the subsequently added corpuscles. 



If we employ the serum of children that have recently recovered 

 from whooping-cough in such a test as this, the result is very demon- 

 strative. In our first experiment we employed the serum of three 

 different children who had recovered from two weeks to a month 

 previously, and as controls the sera of three normal individuals. 



These different sera were heated to 56 degrees and were mixed 

 in doses varying from 0.1 to 0.3 of a cubic centimeter with 0.05 or 

 0.01 of a cubic centimeter of fresh human or guinea-pig serum (alexin) 

 plus 0.2 of a cubic centimeter of an emulsion of the whooping-cough 

 bacillus (a 24-hour culture suspended in salt solution).* Four 

 hours later, after remaining at room temperature, a little well-sen- 

 sitized goat blood was added to each tube. Hemolysis took place 

 in a few moments in the tubes containing normal serum, but the cor- 

 puscles remained intact, even on the following day, in those which 

 contained whooping-cough serum. The sensitizing property, then, 

 of the serum from a case of whooping-cough is very energetic even 

 in so small a dose as 0.1 of a cubic centimeter. It is scarcely 

 necessary to mention that the serum of such cases does not fix 

 the alexin unless the whooping-cough bacillus is present; the other 

 necessary controls are also made. 



The organism resembling the influenza bacillus acts no differ- 

 ently with whooping-cough serum than it does with normal serum. 

 Such an experiment shows not only that this latter organism has 

 nothing to do with whooping-cough, but serves to separate the two 

 bacterial species from one another, f In other experiments the 

 same results were obtained. We tested the sera from two different 



* The experiment may be done with either of these alexins, but the guinea-pig 

 alexin is, in general, more favorable for hemolysis and therefore better to use. 



t It should be noted that the influenza bacillus has in itself a certain property 

 for absorbing alexin without the presence of any sensitizer. We have already 

 stated that the whooping-cough bacillus does this only when whooping-cough 

 serum is present. 



