132 FRANK C. BECHT AND JAMES R. GREER 



the concentration was the same in both. The fact that the concen- 

 tration of agglutinins may be greater in the thoracic lymph than in 

 the serum, renders it hard to see how these antibodies can come from 

 the blood by pure nitration, for in that case, we should expect the 

 hemolysins to run a parallel course a thing which they do not do 

 or else we must assume that the agglutinins pass through membranes 

 more readily than the hemolysins. It would be necessary, also, on 

 the basis of filtration, to assume sudden great changes in the concen- 

 tration of the agglutinins in the blood, for on no other basis could we 

 explain the fact that the concentration of agglutinins would be so 

 much lower in the serum by the time the lymph reached the upper 

 end of the thoracic duct, than it was at the time the lymph was 

 formed. Of course other explanations are possible: there may be 

 an active secretion of the agglutinins into the lymph from the blood, 

 or the agglutinins, after being formed in the area drained by the 

 thoracic duct, are thrown into the lymph, reaching the blood by that 

 route. Much more investigation must be made before any conclusion 

 can be reached on this point. 



The pericardial fluid when collected under the best conditions 

 never shows hemolysins for rabbit corpuscles. Agglutinins may or 

 may not be present. In four of our ten supposedly normal dogs 

 hemolysis was noted, in only one case amounting to more than 10 per 

 cent. Of these four animals, two were in poor condition, emaciated, 

 and generally run down, and both these dogs yielded excessive 

 amounts of pericardial fluid; in the other two cases, the pericardial 

 was found to contain a few erythrocytes. Agglutinins were found 

 in all four of these cases and in three others, making a total of 

 seven in ten. From these experiments we are inclined to believe that 

 hemolysins are not found in the pericardial fluid of normal dogs. The 

 fact that some animals showed hemolysins in the pericardial fluid 

 we would explain as a pericardial transudate in two cases, and to 

 admixture with blood in two cases. We did not test whether it was 

 amboceptor, or complement, or both which was absent from the 

 fluid, altho we have evidence on this point in immune animals. 

 Agglutinins for rabbit corpuscles may or may not be present in the 

 pericardial fluid of normal dogs. 



As will be seen from Table i the cerebrospinal fluid and aqueous 



