366 Dr) ©. W. Barratt. [ Dec. 5, 
the cause of this difference in respect of complement fixation. It would be 
expected that the degree of fixation would be likely to be marked when the 
tumour was considerably degenerated. In the earlier experiments the 
condition of the tumour in this respect was not noted. In the last five 
experiments, however, attention was paid to the condition of the tumour, and 
deviation was found to be increased when marked degeneration was present 
(two cases), but was unaltered when degeneration was slight in rapidly 
growing tumours (three cases). 
Expervments in which the Complement envployed was that Present in the Serum 
of the Mouse. 
Exception may be taken to the experiments already described on the 
eround that there is no direct proof that the compiement employed fits the 
amboceptor of the autotoxin for which search has been made. There is, 
however, indirect proof that this is so, since the complement employed is 
fixed by the amboceptor in the presence of antigen contained in the tumour 
extract. As, however, the degree of complement fixation observed in the 
preceding experiments was slight, it was decided, in view of the importance 
of the issues involved, to repeat these experiments in a form in which this 
objection was avoided, namely, by employing as complement that present in 
the normal serum of the mouse. 
Method.—The lines along which this investigation proceeded are as 
follows: The first stage was to discover a suitable source of red blood cells, 
preferably one for which the serum of the mouse normally possessed a 
hemolysin. The action of the normal serum of the mouse upon several 
easily obtainable varieties of red blood cells was tested, with the results 
given in Table V. The red cells employed were those of the rabbit, guinea- 
pig, rat, man, and goat. It will be seen that the degree of hemolysis 
obtainable with these red cells was much less than that ordinarily occurring 
when a hemolytic serum prepared by injecting red cells intraperitoneally is 
employed. When the different varieties of red cells are compared it is seen 
that those of the rabbit were hemolysed far more readily than any of the 
other forms employed. The red blood cells of the rabbit were therefore 
selected as the most convenient to employ for testing for deviation of the 
complement of the mouse. 
The hemolysin in question in normal serum of the mouse is present in 
relatively small amount. Subsequent observations showed that the amount 
normally present was occasionally considerably less than that indicated in 
Table Y. 
