Water for Bacteriological Analysis. 95 
that the stopper is not jammed in the neck of the bottle 
from unequal expansion in the hot air sterilizer. 
In working at considerable depths I have found it con- 
venient to use a screw at O. This increased pressure upon 
the wings of the clamps holds the stopper more firmly. At 
other times the screw is not needed. By substituting a 
wire for the string attached to the cross bar, the opening 
and shutting can be readily controlled at very great depth. 
A bottle is removed by simply compressing the wings of 
the clamps and lifting it out from the jaws. The ease and 
rapidity with which the apparatus works will be understood 
from the fact that [ am able to collect 10 separate samples 
of water at a depth of 20 feet in from i0 to 15 minutes. 
The bottles made use of are those dropping bottles 
fitted with ground glass pipettes now in common 
use for holding histological reagents. Both ends of the 
pipettes are sealed up in a gas flame, thus converting 
them practically into glass rods. As these bottles are kept 
in stock in the laboratory, one can always be replaced 
if it happens to be broken. The ones I employ hold 50 
e.cm., but I would have preferred 100 c. cm. bottles had 
they been obtainable. The method of clasping the bottle 
by the neck admits of various sizes being employed in the 
same frame as there is space to spare between the cross 
bars. 
The differences between the model here described and the 
' original form introduced by Dr. Ellis are that the bottle is 
grasped by the neck instead of being forced into a socket 
from above. The use of spring clamps to hold the bottle, 
enables bottle and stopper to be brought into position by a 
single act instead of taking them apart and putting them in 
separately. The chief advantage of using the dropping 
bottles described lies in its giving a long tapering stopper, 
the lower end of which remains in the neck of the bottle 
when open, and guides it back into position, and it seemed 
preferable to use a bottle readily obtainable rather than to 
order a special form, which could not be replaced if broken. 
The little sinking frame 1 have just described was ori- 
