150 Scientific Proceedings (67). 



that whole rabbit blood plus immune serum killed the pneumococci. 

 Therefore rabbit blood was taken in hirudin; varying amounts 

 of immune serum were added to the hirudin blood; definite quan- 

 tities of bacteria were added and plates made immediately after 

 starting the experiments, and at intervals for twenty-four hours. 

 The colonies were found much reduced in the tubes containing 

 the immune serum and whole blood, but after twenty-four hours 

 all tubes were " saturated " with bacteria. It was believed that the 

 immune serum plus whole blood produced great reduction in the 

 number of bacteria, but could not cause sterilization. 



Later it was deemed advisable to follow with the microscope 

 the processes occurring in the tubes. To our surprise, we found 

 that the bacteria were agglutinated in tubes containing serum in 

 a dilution of 1-500; macroscopically the agglutination titre is 

 I-80. Thus it appeared that the reduction of colonies was due 

 to the clumping. 



It was then surmised that the disappearance of the bacteria 

 from the circulating blood of the rabbit following the injection of 

 immune serum might be due to clumping in vivo and filtration 

 by the capillary systems of the organs. Our investigation of this 

 point gave the following results: 



When a rabbit with pneumococcic septicemia is given 1 c.c. of 

 immune serum intravenously, the cocci are clumped within forty 

 seconds' time and after two minutes they have left the circulation. 

 We have found the clumps in the heart's blood. Next, fragments 

 of the organs — lungs, spleen, liver, kidney, brain, etc. — were 

 crushed and examined, and clumps of pneumococci were found in 

 all. The fate of the clumps was then investigated. By killing 

 the animals at various times after the administration of the serum, 

 it was observed that the polymorphonuclear leucocytes englobed 

 and digested them. The fixed cells play a small part also. Sec- 

 tioned and crushed tissues gave the same results. Pneumococci 

 from 150 c.c. of bouillon are thus destroyed within two to three 

 hours. The smallest amount of serum that will influence the 

 infection causes the clumping in vivo. 



Typhoid bacilli, dysentery bacilli, streptococci, staphylococci, 

 and gonococci have been tested for agglutination in vivo and all 

 behave in a manner similar to the pneumococci, that is, they 



