96 BACTERIOLOGICAL DIAGNOSIS. 
tube, lay the pipette on its side, and make a mark with 
ink or with a grease pencil to show how high the serum 
reaches. Now suck the column of fluid a little way into 
the tube, and insert the tip of the pipette into the emulsion ; 
suck the latter up the tube until it reaches to the mark. 
This will give you the same amount of emulsion as of 
blood serum; the two fluids will be separated by a short 
column of air. Now withdraw the tip for a moment 
and suck up another small quantity of air; dip it into 
the emulsion and suck it up to the mark again. This 
will give you twice the amount of emulsion as of 
blood serum, the three portions of fluid being separated 
by air (fig. 18). Repeat this process until you have 
nine times the amount of emulsion as of serum; then 
suck the fluid still further from the tip of the pipette, 
and seal the latter ina flame. Probably by this time 
the fluids will have mixed together; if not, tap the 
tube gently (holding it upright) until the air-bubbles 
which you have sucked up make their escape and the 
fluid forms a continuous column. 
Fill another tube to a similar height with emulsion 
(for a control) and place the two side by side in an 
upright position for twelve hours. 
Now examine the fluid in the pipettes. If the emul- 
sion has been made from a living culture the control 
pipette (7.e., that to which no blood has been added) 
will probably remain turbid; if an emulsion of dead 
bacilli has been used it will have become clear, and the 
bacilli will form a uniform even layer at the bottom of 
the pipette. 
Compare the control tube with the pipette to which 
blood has been added. If the reaction is negative 
the appearances will be exactly the same in each; 
but if the reaction is positive the bacilli will fall to the 
