158 
equal to that of the aggregate length of all the water drops in E at any 
one time. But the air escapes through H, under control of a pinchcock, 
and the water is forced out through G. The waste water escapes through J. 
The flow of air through H must be so regulated that water is forced 
out through G just a little faster than it enters from F. This provides for 
an occasional release of surplus pressure by the escape of air through G 
and prevents the filling of K with water, as will be the case if the air is 
allowed to escape too fast through H. The only irregularity of flow is at 
the time of the release of pressure through G; but the air stream is seldom 
interrupted for more than a few seconds, and by careful adjustment the 
frequency of these interruptions may be reduced to a minimum. Perfect 
adjustment would entirely eliminate these irregularities by allowing the 
water to escape through G just as fast as it enters through F; but, per- 
fection being impossible, it is better to have the interruption occur as an 
escape of air through G than of water through H. 
Theoretically the pressure of the air issuing from H, and consequently 
the depth to which a solution can be aerated, is determined by the vertical 
distance from the level of the water in K to the outlet of G. In practice, 
however, the apparatus fal!s somewhat short of this, due to friction of 
the air through H and the capillarity of the liquid to be aerated. The 
density of the culture solution is, of course, a determining factor also. 
The efficiency of the apparatus depends largely upon the nature of the 
tube F. If it is of too small bore, the friction is too great; and if it is 
too large, the water has a tendency merely to run down the inside surface 
and fails to carry any air with it. <A very satisfactory size of tube is 
one haying an internal diameter of 2 to 4 mm. If a larger quantity of air 
is needed at H, tne pressure to remain the same, it is better to use two 
tubes for F than to try to increase the capacity by substituting one larger 
tube. If the pressure is to be increased and the amount of air to be de- 
livered in a given time is to remain the same, G must be lengthened, and 
this may necessitate the lengthening of F also, for IF will carry air only 
so long as the aggregate length of its water column is greater than that 
in G. In adjusting the apparatus, glass or metal stopcocks have been 
found more satisfactory where the flow of water is to be regulated, while 
pinchcocks on pieces of rubber tubing have been found best for regulating 
the stream of air. 
When well adjusted and in good working condition the apparatus is 
economical. Tests on the one now in use have shown that it can be made 
