WIDAL’S METHOD. 87 
avoid air-bubbles. This is because the test is easier to 
carry out and is more conclusive if clear serum can be 
obtained. The test may be applied to a dry drop of 
blood, but in this case there is no possibility of making 
an accurate dilution. The test is wonderful enough as 
it is, but the practitioner should not expect too much 
from it if he sends a few splashes of blood inside a 
capillary tube of the pipette. This is so important that 
we repeat: do not be satisfied unless the bulb is com- 
pletely filled with blood. When this is accomplished 
suck the fluid a little further up and seal the pointed 
end of the pipette in a flame—the flame of a wax match 
will answer. 
If a pipette is not at hand it is advisable to manufac- 
ture one out of any piece of glass tubing which may be 
available. Failing this the best plan is to dry several 
large drops of blood on a clean slide; but results ob- 
tained from blood forwarded to a laboratory in this way 
should be received with caution, for the dilution is 
necessarily a matter of guess-work. 
Do not send blood in vaccine tubes. When this is 
done the blood often adheres so firmly to the glass that 
the whole tube has to be ground down in a mortar with 
a drop or two of water. 
METHOD OF PERFORMING W1DaAL’s REACTION BY THE 
Microscoric METHOD. 
Requisites—1. A young culture (not more than eigh- 
teen hours old) of typhoid bacilli on agar. 
Where dead bacilli are to be used this is to be re- 
placed by a culture obtained from the laboratory and 
prepared in the manner described. 
