THE DIPHTHERIA BACILLUS 431 
proportion of non-specific protein as far as is practical by })urely phy- 
sical agents. This reduction of non-specific proteins is important for 
two reasons: (1), Because it reduces the danger of anaphylaxis due to 
sensitization of the patient; and, (2), because the rashes and joint 
swelhngs are notably reduced when the concentrated antitoxin is used 
instead of the whole horse serum. It is possible to obtain the same 
therapeutic effect by the injection of about one-third the amount of 
solution when concentrated antitoxin is administered. 
Properties.— D\^\ii\\Qv\2L antitoxin specifically neutralizes diphtheria 
toxin both iji vitro and in vivo. It has little neutralizing value for 
the toxone, however; consequently in severe cases when it is used late, 
it will not prevent the development of postdiphtheritic paralyses. It 
has both prophylactic and curative properties. It is not bacteriolytic 
and exhibits no agglutinins for diphtheria bacilli. Nothing is definitely 
known of the natiu-e of diphtheria antitoxin. If the diphtheria toxin 
is a ferment, the antitoxin would appear to be an antiferment. The 
fact that it is precipitated with the globulin fraction of the blood serum 
would suggest that it may be either closely related to the proteins or 
a true protein itself. 
Di])htheria toxin varies considerably in its potency due to the fact 
that it deteriorates; the antitoxin, on the contrary, is more stable. 
Consequently for purposes of comparison and standardization a 
standard antitoxin is used. Two such standard antitoxins are recog- 
nized officially: one prepared in the Ehrlich Laboratory in Germany; 
the other prepared by the United States Public Health Laboratory in 
Washington, D. C. Both of these antitoxins are prepared on a very 
large scale and preserved in a cold, dark, dry place in packages of con- 
venient size. When the supply of one or the other of the standards is 
nearly exhausted a new lot of antitoxin will be prepared and carefully 
compared with the old. Small amounts of the standard antitoxin con- 
taining a definite number of antitoxin units are sent out regularly by 
the central laboratories to interested laboratories for testing purposes. 
Standardization of A7ititoxin. —The antitoxin unit may be defined 
as ''that amount of antitoxin which just suffices to protect a guinea- 
pig of 250 gm. weight against one hundred times the minimal fatal dose 
of diphtheria toxin." The process of standardization of antitoxin of 
unknown potency is carried out in the following manner: diphtheria 
toxin, prepared as described above, is mixed in gradually diminishing 
amounts with a definite amount of the stanflard antitoxin (containing 
a known number of antitoxic imits) and allowed to stand for twenty 
to thirty minutes to permit uni(m of the toxin-antitoxin to take place. 
The mixtures are then injected subcutaneously into guinea-pigs of 
250 gm. weight. The greatest dilution of toxin which kills a guinea- 
pig in four days is said to be the L + dose — that amount of toxin 
which will neutralize (say) 100 antitoxin imits and leave an excess of 
toxin just sufficient to kill the animal. Having found the L+ dose 
of toxin (which standardizes its toxicity in terms of standard anti- 
