Types of Pneumococci 477 



mined by puncture of the abdominal wall by a capillary pipet, the cocci will have 

 increased sufficiently and the mouse can be chloroformed to death. The abdomi- 

 nal cavity is then Carefully opened and the fluid in the peritoneal cavity trans- 

 ferred with a pipit to 4-5 cc. of physiological salt solution in a centrifuge tube. 

 After mixing, the contents are centrifuged for a minute or two at low speed to 

 throw down leukocytes, shreds of fibrin and any clumps of bacteria, and the 

 supernatant fluid which is a fairly uniform, slightly opalescent suspension of 

 pneumococcic removed with a pipet to a second tube and centrifuged at high 

 speed to thrown down the cocci. The supernatant fluid is thrown away, and the 

 bacterial sediment distributed throughout fresh salt solution and placed in 

 several small tubes. To one of these, a small quantity of the serum agglutinating 

 type I, to the other a small quantity of that of type II, to a third that of type III 

 is added, and the tubes stood in an incubating oven for eighteen to twenty-four 

 hours. If one agglutinates, the type is at once known; if none agglutinates, the 

 disease affecting the patient is caused by type IV. 



■ The type organism now being known, the therapeutic antipneumo- 

 coccic serum appropriate in antagonizing type I or type II should 

 be administered. If the disease is caused by type III, the patient 

 cannot be benefited, as no serum has thus far been produced that 

 is of use in treating cases of pneumonia occasioned by type III. 

 A second method of determining the type of pneumococci in any 

 case has been suggested by Blake* and depends upon specific 

 precipitation. Blake points out that the method by agglutination 

 is sometimes delayed by the occurrence of complicating infections 

 — influenza bacillus, mouse typhoid bacillus — and is always delayed 

 by the time of incubation necessary for the occurrence of the 

 agglutination. 



To obviate these, he proceeds according to the method given above, inoculating 

 the mouse with the sputum, making occasional punctures with capillary pipets 

 to determine when the organism,s have multiplied sufficiently, kills the mouse 

 at the appropriate time, collects the fluid from its abdominal cavity, dilutes it 

 with 4-5 cc. of salt solution and immediately centrifjuges at high speed to throw 

 out the leukocytes, fibrin and bacteria, leaving a clear fluid. This is saved and 

 constitutes the fluid susceptible of specific precipitation by the immune serums. 

 Five-tenths cubic centimeter of the fluid is placed in each of a series of small 

 ' test-tubes, and to each 0.5 cc. of the immune serums of types I, II, III, and IV are 

 added. An immediate precipitate takes place in that tube containing the homo- 

 logous immune serum, the other tubes remaining clear. The association of the 

 micro-organisms mentioned above does not interfere with the success of the 

 specific precipitation test. 



Blake, and also Stillmanf both found that the organisms of type 

 II were not always uniform in behavior and have therefore divided 

 them into several sub-types Ila, IK and IIx. 



Stillman observed and studied 454 cases of pneumonia and found 

 the organisms as follows: 



Type 1 151 33.26 per cent. 



Typell 133 29.29 per cent. 



Type Ila 6 1.32 per cent. 



T3rpe lift 4. 0.88 per cent. 



T3^e Ila; 9 .1.98 per cent. 



Tjrpe III 59 12.99 per cent. 



Type IV 92 20.26 per cent. 



* li 



'Jour. Exp. Med.," 191 7, xxvi. No. i, p. 67. 

 t "Jour. Exp. Med.," 1917, xxvi. No. 4, p. 513. 



