Juxy 12, 1918] 
washing is necessary. Nearly everywhere the 
pressure in water pipes varies during the day 
and the night. Either the dripping may be- 
come so rapid as to flood the funnel and cause 
a loss of the organisms or the flow may cease 
altogether, leaving the material in an exposed 
condition and hence subject to drying. To 
overcome this trouble the following device was 
used. 
Prepare a funnel and filter paper in the or- 
dinary way. Then fill a two or three gallon 
bottle with distilled water and stopper it. The 
stopper should be a one-holed rubber one 
through which passes a glass tube of desired 
length but having an inside diameter of at 
least one quarter of an inch. Make the stopper 
very secure. Empty the organisms previously 
fixed, into the filter and then quickly invert 
the two or three gallon bottle; allow the tube to 
extend into the funnel so that the end of it is 
about a half inch below the edge of the filter 
paper. Give a good support to the reservoir. 
See the accompanying diagram for arrange- 
ment. 
As filtration goes on the surface of the water 
in the funnel falls below the end of the tube 
that projects into it. This allows air to pass 
into the tube; the ascent of the air causes 
water to flow out of the tube and replenish the 
supply in the funnel. The flow continues till 
the surface of the water reaches the tip of the 
tube when the supply is automatically shut off. 
The process is continued in this way as long 
as there is water remaining in the reservoir. 
The descent of the water causes suflicient dis- 
turbance of the material in the funnel to 
prevent the organisms from matting against 
the side of the filter paper. 
The rate of filtration and consequently the 
flow of water from the reservoir may be regu- 
lated by using various grades of filter paper or 
several sheets of a thin quality. The writer 
has found that a rather heavy grade of paper 
permits sufficiently slow filtration so that three 
gallons of water will last about fourteen con- 
tinuous hours. 
Herpert RuCKES 
ZOOLOGICAL LABORATORY, e 
Grove City COLLEGE 
SCIENCE 
45 
AN OPTICAL ILLUSION WITH FATAL 
CONSEQUENCES 
THERE is an optical illusion that has prob- 
ably led, within comparatively recent times, to 
the death by drowning of scores, or even hun- 
dreds, of capable but inexperienced swimmers. 
A person swimming with the wind, and con- 
sequently with the waves which travel in the 
same direction faster than it is possible to 
swim, receives the impression of being carried 
backward by the water. In the absence of 
knowledge or information covering the case, 
most persons, so situated, if headed toward the 
shore, immediately think of “undertow,” a 
word which nearly every one has heard, and 
believe themselves to be caught in an offshore 
current. The instincts of an untrained or 
half-trained swimmer always lead to a nervous 
haste and overexertion in deep water, even 
under conditions most favorable for swim- 
ming. When these instincts are supplemented 
by the panic that arises from the belief that the 
person is caught in an “undertow,” the re- 
sulting increase of effort and acceleration of 
action reduces efficiency to a degree that must 
certainly have left many persons fatally ex- 
hausted before they reached a footing. 
My attention was first called to this phenom- 
enon through two cases of able-bodied but in- 
different swimmers who, after swimming just 
beyond their depths in an onshore breeze at 
Pasay Beach, near Manila returned to the 
bathhouse in an excited state and reported 
having been caught in an “undertow” with 
nearly fatal result. In each case I made im- 
mediate investigation of the water at the point 
indicated and found neither “undertow” nor 
offshore current sufficient to embarrass any 
swimmer. Subsequently, on numerous 0c- 
casions, while initiating beginners into deep- 
water swiniming, being headed for shore with 
an. onshore breeze, I have heard the initiate re- 
mark, with deep concern, that there was a 
current against us. This required to be ac- 
counted for. The feeling of being carried 
backward may be satisfactorily explained to 
most persons as arising in the same way as 
the effect commonly produced on a person 
