493 
been injected in each instance, then the serum in the individual doses was as 
follows: 0.1, 0.05, 0.025, 0.01, 0.005, 0.0025, and 0.001. The result was taken 
after twenty-four hours. 
It is necessary to remember that dysentery bacilli very rapidly lose 
their virulence, so that, after some weeks’ standing they are already not 
appropriate for the purpose of testing sera; however, I have succeeded 
in maintaining their virulence successfully during more than one year 
by placing the cultures of the strains in the refrigerator immediately upon 
their isolation. When they are to be used some of this culture is retrans- 
planted to agar. 
A few examples of my numerous experiments with different sera are 
given below. (See Table VI.) The third type of the bacillus is lacking 
in this table, as the cultures which I had at my disposition had already 
lost their virulence and were therefore no longer available for animal 
experiments, but I have shown in other instances that Type III is acted 
upon almost as well by serum from Type IT as is the latter bacillus itself, 
As the table demonstrates, the serum from Type I more or less affects 
Types IV and V, that from Type IL feebly acts upon Type I, and is 
almost without effect upon Types IV and V, but, on the other hand, 
Types IV and V are closely related to each other, so that a serum from 
one of these acts almost as well upon the other as it does upon itself. 
The above experiments show that a universal serum can be obtained 
by mixing the immune serum secured from Types I and II with that 
from ITand IV. It is true that if we take three horses and immunize one 
of them against Type I, the second against Type II, and the third 
against ‘Type IV, and then mix the sera prepared from all three, we 
also would obtain a universal serum, but the disadvantages of this 
method would consist in the fact that the individual titer of each serum 
would be reduced to one-third of the value which it would have if it alone 
were present. 
For the purpose of immunizing horses the alternating separate injec- 
tion of cultures of two types is recommended. If a mixture of the two 
is used, serious suppuration at the point of injection generally takes 
place, with the result of greatly reducing the acquired immunity. It is 
also true that as high an immunity is not obtainable if the two types are 
kept s¢parate but injected simultaneously on both sides of the animal 
as if they were given in alternate doses, because then the quantity of any 
one used would only be one-half of the total. It must also be borne in 
mind that horses are very susceptible toward the original type, whereas 
they can very well endure the varieties; therefore the immunization can 
be forced about twice as far with the latter as it can with the former. 
Owing to these considerations my advice is as follows: To immunize 
two horses simultaneously, the one with Types I and II, the other with 
I and TV, as soon as a high immunity is reached, serum should be taken 
from both and mixed in equal quantities. By this means the best 
universal serum can be obtained. 
