PART 3.— THE TOXIC PRINCIPLE. 
THE EFFECT OF HEAT UPON THE TOXICITY OF HORSE 
SERUM. 
It will be seen from the following tables that normal horse serum 
may be heated to 90° C. for one hour and still remain slightly toxic 
when injected into a sensitized guinea pig. Its toxicity, however, 
is evidently markedly affected. Heating to 70° C. for one hour does 
not seem to appreciabl}^ diminish its poisonous properties, but it 
appears to be affected at 80° for one hour. At 100° for one hour the 
toxicity apparently disappears. 
Blood serum of course can not be heated to these high degrees 
without coagulation, and it is therefore necessary to dilute it in the 
proportion of one part of blood serum to three parts of distilled water. 
This dilution may then be heated to a high temperature without 
producing any visible change other than a slight opalescence. 
It appears that there is a slight difference between the sensitizing 
and toxic principles in horse serum so far as resistance to heat is 
concerned. Serum heated to 100° C. for one hour retains some power 
of sensitization, but seems to lose its toxicity when given at the 
second injection. This difference maybe more apparent than real, 
for exceedingly minute amounts are sufficient to sensitize guinea pigs, 
while a very large quantity of weakened serum would be necessary 
to produce symptoms. It must be remembered that in our experi- 
ments 20 c. c. of the dilution represents but 5 c. c. of the serum. 
These facts must be considered in drawing conclusions from work 
upon split proteins, fractional precipitation, or other methods to 
isolate the sensitizing substance in pure form. A very minute 
amount of the original protein substance in horse serum clinging to 
the globulins or other substances modified by chemical methods 
might be sufficient to sensitize guinea pigs, whereas it would require 
very large amounts of such a modified protein to poison a sensitive 
animal. 
( 15 ) 
