544 On the Presence of Hem-agglutinins, etc., in the Blood. 
culture from the hemolytic tubes and from the blood at the post-mortem 
examination, was an atypical streptococcus, according to the classification of 
Andrewes and Horder. The micro-organism itself had also the power of 
exciting hemolysis, so that the hemolytic action which was found to take 
place in this sample of immune serum was not due to the serum itself, but to 
the streptococci which it contained in such large numbers. 
The third example was a severe case of this disease in which there was no 
auto-hemolysis, although the blood was examined on several occasions, and 
there was no auto-hemolysis even after the immune serum had been com- 
plemented, but a high degree of hemolysis occurred by allowing the immune 
serum to act on the red blood corpuscles from a case of acute lobar pneu- 
monia; 75 per cent. of serum produced complete hemolysis ; 62°5 per cent. 
and 50 per cent. of serum produced incomplete hemolysis; 37°5 per cent. 
slight, and 25 per cent. very slight, hemolysis. 
In a case of very severe anemia, due to hemorrhage, there was no 
hemolysis, but the immune serum had the power to a slight degree of 
hemolysing normal red cells. The second case was an example of acute 
lymphocythemia in which there was no auto-hemolysis, but a slight 
hemolytic action of the immune serum on normal red cells. 
Group VII.—In the streptococcal cases hemolysis occurred in two 
instances. The first was a.case of streptococcal pyzemia due to an atypical 
streptococcus, in which the immune serum produced well-marked hemolysis 
on normal red cells down to a dilution of 25 per cent. of serum, while there 
was no auto-hemolysis. The second case was an example of acute erysipelas 
due to the Streptococcus pyogenes, in which only very slight hemolysis 
occurred in the presence of 75 per cent. of serum. Here, again, the inter- 
action was limited to the immune serum and the normal red cells. 
Group IX.—In the case of chronic lead poisoning there was no auto- 
hemolysis, but the immune serum had a slight but distinct hemolytic action 
on normal red cells. 
In a case of jaundice, probably carcinomatous, there was distinct auto- 
hemolysis with 75 per cent., 62°5 per cent., and 50 per cent. of serum. This 
case, however, was the only example in the whole series of experiments in 
which there was an auto-hemolytic action. If this case proved to be 
carcinomatous, it would be the only example of this disease in the whole 
series investigated. 
Conclusions. 
It will be noticed that a striking feature, with one exception, in all these 
experiments is the absence of the auto-hemolytic action and the hemolysing 
property of the same serum on normal red blood corpuscles. 
