136 FRANK C. BECHT AND JAMES R. GREER 



hemolysis where only traces had appeared, and to produce hemolysis 

 in. other cases where there were no traces with exactly the same 

 amount of the fluid without complement. 



The addition of c6mplement in a non-hemolytic dose is able to 

 cause hemolysis in the pericardial fluid. In 8 of 13 experiments on 

 blood-immune dogs traces of hemolysis occurred in the lowest dilu- 

 tions of that fluid. In these experiments we could detect no contami- 

 nation with blood, neither did we note that the fluid was present in 

 excessive amounts. Apparently, then, in dogs immune to a foreign 

 blood, amboceptor is always, and complement usually present in the 

 pericardial fluid. This agrees in part with the findings of Mioni 

 who found amboceptor but no complement for guinea-pig corpuscles 

 in the pericardial fluid of the ox. No amboceptors are found in the 

 cerebrospinal fluid or aqueous humor, for no hemolysis occurs when 

 effective complement is added, at least a complement which proved 

 effective in the case of the other body fluids. 



The agglutinins in the various body fluids are seen from Table 5 

 to run practically the same course as in the normal animal, except 

 that the^concentration is highest in the serum, a little lower in the 

 thoracic lymph, still lower in the neck lymph, and lowest, but always 

 present, in the pericardial fluid. Sometimes, as in the normal animal, 

 the concentration of agglutinins in the thoracic lymph is equal to or 

 greater than that of the serum. In the 16 immune animals in which 

 we compared serum and thoracic lymph, in n cases the serum was 

 higher in the concentration of the agglutinins than the thoracic 

 lymph, in two cases the thoracic lymph was higher than the serum, 

 and in three cases the two fluids showed equal concentration. The 

 fact that the thoracic lymph may contain these antibodies in equal or 

 even considerably higher concentration than the corresponding serum 

 may be of significance as bearing upon the source of these substances. 



In 15 experiments with cerebrospinal fluid from immune dogs, 

 a positive agglutination was secured in two. In 16 experiments with 

 the aqueous humor, positive results were secured in five. Thus it 

 will be seen that agglutinins may be found in the cerebrospinal fluid 

 and the aqueous humor, but their presence is the exception and not 

 the rule. 



A point of considerable interest is the determination whether the 



