152 C. J. MARTIN. 
heated to 85° C. it only possesses this power to a trifling extent, 
and to compass the death of the animal by intravascular clotting, 
the dose must be increased five-hundred fold. 
An obvious inference from these facts is that the body in 
venoms which destroys blood cells and kidney epithelium is the 
proteid coagulated by heat. 
It must however be borne in mind that heat affects venoms in 
two ways :— 
_ (1) By coagulating some of the proteids present, in which con- 
dition they are inert. 
(2) By imparing the toxic power of the proteids present without 
influencing their solubilities, or indeed changing them in 
any way recognisable by chemical tests. 
The first method of action is sudden. When the solution is 
raised to a certain temperature some portion of the proteid con- 
stituents is coagulated. 
The second method is gradual, the longer. the heating and the 
greater the dilution of the solution the more impairment of viru- 
lence occurs. Perfectly dry venoms may be submitted to a 
temperature above 100° C. without diminishing their toxic power. 
Cobra poison is least affected by heating, whereas viperine poisons 
are very sensitive to heat when in solution. 
Accordingly, before I could conclude that the alteration in the 
action of Pseudechis venom was due to the exclusion of one of the 
poisonous constituents by heat, it was necessary to accomplish 
the separation by some other means which should not at the same 
time modify the remaining constituent. This I have been able 
to do, and by a method which has thrown light upon the causation 
of those differences in effects which follow subcutaneous or intra- 
venous inoculation. The method consists in filtering a solution 
of venom through a film of gelatin by aid of a hydrostatic pressure 
of fifty atmospheres.'_ The proteid which is coagulated by heat 
1C. J. Martin— Note on a method of separating Colloids from 
Crystalloids by Filtration.—Proc. Roy. Soc. N.S. Wales, 1896 and Journ. 
of Physiol. 1 1896, 
