42 
sticking to each other. As it is probable that some of these bottles 
will be kept for }"ears, this may be an important detail. The neck of 
the bottle is carefully dried and then quickly dipped into melted paraffin 
with a high melting point so as to seal the joint completely and prevent 
evaporation or contamination. 
Each bottle is labeled and stored in a special ice box kept at a con- 
stant temperature of 5° C. 
It will be seen that according to this method the toxine is kept in 
absolute darkness, free from contact with the oxygen of the air, and 
at an equable and low temperature. By the use of the ordinaiy bac- 
teriological precautions the addition of a preservative as toluol and 
similar substances is rendered unnecessaiy. 
TESTING THE TOXINE. 
It is first necessaiy to determine the minimal lethal dose of the tox- 
ine with approximate accuracy. W eak poisons are discarded. A fresh 
toxine requiring more than 0.01 c. c. to kill a 250-gram guinea pig is 
too weak for the purposes of this special work. 
If the preliminaiy tests show that the poison has the required tox- 
icity, its minimal lethal dose must then be accurate!}" determined. 
This may require a number of animals. 
After the absolute toxicity of the poison has been determined it is 
necessary to establish the L+ and the L® doses. An illustration of 
how these doses are obtained follows. These results are taken from 
our records. 
Tests to determine the L~r dose of toxine No. 7. 
Result. 
1 immunity unit —0.19 c. c. toxine 
1 immunity unit -|- .20 c. c. toxine 
1 immunity unit — .21 c. c. toxine 
1 immunity unit — .22 c. c. toxine 
Invariably causes late paralysis, never 
acute death. 
Sometimes causes late paralysis and 
sometimes acute death. . 
Always causes acute death about the 
fourth day. 
Always causes acute death, usually on 
the second or third day. 
The L -j- dose of this toxine is, therefore, just 0.21 c. c. 
