Haemolysis hy Serum in Combination ivith Certain Benzol Bodies. 61 
It has already been noted how the haemolytic action of the serum with 
brilliant green is not due to complement, since serum heated to 55° C. is still 
capable of acting. In order to ascertain whether this property was stable 
at higher temperature, tests were made with diluted rabbit's serum which 
has been boiled for two minutes. It was found that boiled serum was as 
active as fresh serum (Table YIII). 
Rabbit's serum was also fractioned into (1) Albumin, (2) pseudo- 
globulin from " end-piece," and (3) " mid-piece " (by method of Browning 
and Mackie) , and these fractions were tested separately and in combination 
with the treated corpuscles. As regards the haemolytic property of the 
serum, no qualitative fractioning occurred ; all these protein fractions 
separately exerted a certain degree of haemolytic action which was less than 
that of the whole serum. When combined the full c|uantitative effect was 
reproduced, but only by " summation of effects " (Table IX). 
Lecithin was also found to produce haemolysis of corpuscles treated with 
brilliant green (Table X). 
That the haemolytic action of serum with brilliant green was not due to 
the action of lipoids (e. g. lecithin) present in it was proved by the fact that 
serum which had been extracted with alcohol and ether still retained its 
haemolytic activity (Table XI). Rabbit's serum was treated with a con- 
siderable excess of alcohol ; the precipitated protein was removed by centri- 
fugalisation and washed four times with a mixture of alcohol and ether ; 
the precipitate was then suspended in a volume of salt solution representing 
five times the volume of the original serum. 
Thus one must conclude that the active principle of the serum is an 
extremely stable one, and is resident in and distributed among the different 
proteins of the serum. It has been found possible to abolish the activity of 
the serum by filtering it through a Berkefeld filter. In this connection it is 
to be remembered that various enzymes are similarly retained in porcelain 
filters during filtration (Levy (3)), and, as Muir and Browning (4) 
originally showed, complement is removed from serum by the same process. 
When the sera of a number of different animal species were tested with 
corpuscles of different species, it was found that the action of brilliant green 
in sensitising red corpuscles to the haemolytic effect of the serum was a 
general one, and there was no evidence of specificity as regards the action of 
any particular serum on the corpuscles of different species (Table XII). 
Thus brilliant green is capable of sensitising the corpuscles of an animal to 
the lytic action of its own serum, and the sera of the rabbit, guinea-pig, ox, 
sheep and man are all capable of lysing ox's corpuscles which have been 
treated with brilliant green to a more or less equal extent, though with 
some degree of variation. The lytic dose of sera of different species and of 
different individuals of the same species may vary from 0'005 c.c. to 0*05 c.c. 
The average dose of any serum may be said to be about 0-025 c.c. A 
