VENOM HASMOLYSIS AND VENOM AGGLUTINATION 167 
biological nature of haemolytic and toxic constituents of snake venoms and 
certain toxic serums. In this article Stephens confirmed the peculiar anti- 
hemolytic property of cobra venom when employed in a strong concentra- 
tion, namely, more than 0,cooor gm. in 1 c.c. He failed to find the reason 
for this non-occurrence of haemolytic phenomenon in a concentrated venom 
solution, but at the same time he brought out another puzzling phenomenon 
with cobra hemolysis — that the addition of a large quantity of horse serum 
to such mixture (large quantity of the venom and the blood) induces a rapid 
and complete hemolysis. At that time this was entirely inexplicable, although 
we to-day understand it and I will later return to this phenomenon. 
In testing the neutralizing power of Calmette’s antivenin against the hamo- 
lytic principles contained in various snake venoms, Stephens found that it had a 
very slight neutralizing action on the hamolysins of the venoms of Daboia, 
Crotalus terrificus, and Pseudechis porphyriacus. Against cobra venom the 
antivenin neutralized the lytic action of 0.00045 gm., or about 4 minimal 
lethal doses, while the same quantity could neutralize only 0.00003 gm. of 
daboia venom, a quantity far less than the lethal dose. Stephens states that 
normal or antivenomous serums of horse sometimes produce quite marked 
and progressive haemolysis and sometimes an opposite effect. Thus, when 
antivenin was introduced in a dose insufficient to neutralize cobra venom 
completely, a much prompter hemolysis may occur than in the saline venom 
solution. Here he expresses his uncertainty as to the identity of the hamo- 
lytic principles existing in different venoms. 
Stephens’s observations on the hemolytic and toxic effects of the serums 
of Tropidonotus natrix, python, frog, and eel, as compared with those of 
venoms, are very interesting. With the tropidonotus serum he found a cer- 
tain antihemolytic action of antivenin, while normal horse serum had none. 
He found that a rabbit immunized against the tropidonotus serum to quite a 
high degree succumbed quickly to a single lethal dose of the serum of an 
Australian python, and concluded that the haemolytic as well as toxic constitu- 
ents of these two serums can not be identical. While the serum of horse had 
marked promoting effect upon the cobra as well as daboia venom hemolysis, 
an inhibitory effect was found in the serums of the snake. 
Stephens filtered a 0.25 per cent solution of cobra venom through a 
gelatinized bougie, which had previously been used to filter bullock serum, 
until the filtrate gave no albumin reaction. By repeated filtrations under high 
pressure, the filtrate was tested each time for its haemolytic and toxic actions. 
The filtrate was still active after the third passage, but no more after the 
fourth. A trace of biuret reaction was given with this last filtrate. 
In 1900 Walter Myers contributed another very exhaustive study on the 
nature of toxic constituents of cobra venom. From his own experiments, as 
well as from those of many of his predecessors, notably those of Weir Mitchell 
and Reichert, he concluded a priori that in the venom of cobra di capello 
there are present at least two poisonous substances. One of these is ham- 
olytic, and the other causes death, probably by its action on the respiratory 
