130 FRANK C. BECHT AND JAMES R. GREEK 



Our methods were the following: Quantities of the various body fluids of the animal 

 to be tested varying between o. i c.c. and o.oooi c.c. were placed in a series of eight 

 dry, sterile test tubes plugged as for bacteriological work. In order to make the 

 necessary measurements with a pipette graded to T fa of a c.c., dilutions of the body 

 fluids T V and ^ were made. To the fluid in each tube enough sterile 0.9 per cent 

 NaCl solution was added to make the total volume up to 0.4 c.c. To this was then 

 added o. 2 c.c. of a 5 per cent suspension of the corpuscles to be tested. In this way we 

 got dilutions of the fluid varying between i : 6 and i : 6, 144. All of the fluids from the 

 same animal were prepared, the tubes were placed in a block containing a suitable 

 number of holes, and adjusted to the sliding platform of a shaker in an incubator 

 warmed to 37 C. The shaker was run by water power, and the motion was rapid 

 enough to secure constant, thorough agitation, but not violent enough to injure the 

 corpuscles. The routine technic was to keep the tubes in the shaker for an hour and 

 in the ice-box from 12 to 20 hours to permit sedimentation of the corpuscles before the 

 final reading. 



In determining the amount of hemolysis in the final reading, the following method 

 was employed: A measured sample of the corpuscle suspension in the test was sedi- 

 mented in the centrifuge, and the supernatant liquid drawn off with a pipette. The 

 corpuscles were then laked by adding distilled water to restore the original volume. 

 This sample contained three times as much hemoglobin as the hemolytic tests, because 

 0.2 c.c. of the corpuscles were added to 0.4 c.c. of the fluid tested. Therefore, the 

 above sample was diluted to three times its volume with water. This, then, would give 

 exactly the same concentration of hemoglobin as in any tube in the test, provided that 

 the hemolysis was complete, and is termed 100 per cent for this sample of corpuscles. 

 By further dilution tubes containing 90 per cent, 80 per cent, etc., were prepared. 

 No attempt was made to estimate closer than 10 per cent. A new scale was made for 

 each sample of corpuscles. 



The agglutinins were read from the same tubes as the hemolysins. The method 

 employed to determine whether or not agglutination had occurred, was inspection of 

 the rim of sedimented corpuscles. When the corpuscles, after sedimentation by stand- 

 ing in the ice-box, show a perfectly smooth, knife-edged border, no agglutination has 

 occurred. If the border is slightly or decidedly roughened, agglutination has occurred 

 At first this method was carefully supplemented by microscopic examination, but it 

 was soon found so accurate that in the later experiments we depended entirely upon 

 the observation of the rim of the corpuscles, and dispensed with the use of the micro- 

 scope. 



A. Normal animals. The concentration of lysins and hemag- 

 glutinins in the body fluids of normal animals varies within rather 

 narrow limits. This variation is great enough, however, to make it 

 necessary that the comparison be made between the body fluids of 

 the same animal. The following experiment shows the behavior 

 of the body fluids of the normal dog. 



Table I shows that the concentration of hemolysins is greater 

 in the serum than in the other body fluids; thoracic lymph is next, 

 at least in the case of rabbit corpuscles; and neck lymph is third. 



