CONCENTRATION OF ANTIBODIES 153 



Staph. aureus in the body fluids of normal dogs. The amount found 

 in immune dogs is quite comparable with this, as regards both the 

 amount and the relative distribution. 



Opsonin may be present in considerable concentration in all of 

 the body fluids studied by us, but in the serum, thoracic, and neck 

 lymphs they are always found in greater concentration than in the 

 other fluids studied. Of these fluids the serum usually contains the 

 greatest amount, and the thoracic lymph usually more than the cor- 

 responding neck lymph. The pericardial fluid, cerebrospinal 

 fluid, and the aqueous humor may or may not contain opsonin, but 

 rarely in amounts comparable to those of the serum and lymphs. 



HEMOPSONINS. 



It has long been noted that red corpuscles were taken up by phagocytes under 

 certain conditions but it was not until opsonins were discovered by Wright and Douglas 

 that the role of the serum in phagocytosis ^vvas understood. Neufeld and Topfer33 

 showed that there was something termed by them "hemotropic substance" in the 

 serum of a rabbit immune to goat blood, which caused the phagocytosis of goat erythro- 

 cytes by guinea-pig leukocytes in vitro. Barrats found the same antibody in the serum 

 of doves immune to hen blood. He proved these bodies thermostable. Hektoen 16 

 was the first to point out the similarity of these substances to bacterial opsonins and 

 suggested the name "hemopsonins." Neufeld and Topfer, Barrat, Keith, and Hektoen 

 all cite evidence to show that these substances are distinct from the amboceptors of the 

 hemolysins. The most extensive recent work on the hemopsonins is by Hektoen. ? 

 He shows that normal serum may contain hemopsonins for heterologous or even 

 homologous erythrocytes; that immune hemopsonins are highly resistant to heat; and 

 are in part specific and in part non-specific. 



Methods. The body fluids of the dogs were secured as described earlier in this 

 paper, and inactivated by heat at 55 C. for 30 minutes. Rat corpuscles were used 

 throughout the work. They were made up to 5 per cent suspension after careful 

 washing in 0.9 per cent NaCl solution. The technic followed was to measure into 

 each of a series of small test tubes quantities of fluids ranging between 0.2 c.c. and 

 0.002 c.c. and then adding enough salt solution to make 0.2 c.c. To this was then 

 added 0.4 c.c. of a mixture of equal quantities of rat corpuscles and leukocytic suspen- 

 sion. In this way dilutions varying between 1:3 and 1:300 were secured. The tubes 

 were incubated in a shaker for one hour. Then the contents were mixt thoroughly, 

 smears made, and fixt in absolute alcohol for one hour, and then stained with Giemsa's 

 stain for 30-60 minutes. The percentage was figured from the number of the leuko- 

 cytes actively phagocytic. No attempt was made to count the number of erythrocytes 

 engulfed per leukocyte. In every case 500 leukocytes were counted. Our figures will 

 represent, then, the percentage of leukocytes phagocytic, and shows only the activity of 

 thermostable hemopsonins. 



Early in the work an experiment was performed to test the accuracy of the method. 

 Two. complete sets of tests were made with the fluids of a normal dog, smears were 



