59 
■ of Edinburgh, Session 1882 - 83 . 
and the angle of torsion noted. The pressure in the gasometer was 
then raised, a stronger current was blown over the water, and the 
reading again taken. In this way a number of readings were taken 
at different pressures, the angle of course increasing with the pressure. 
After taking two or three series of such readings to get an average 
for the clean water surface, a little oil was put on the water, and 
another series of readings taken. The result of these experiments was 
that there was no decrease in the amount of deflection after the oil 
was added to the- water ; which shows that as the air communicated 
the same amount of motion to the oily 'surfaced water as it did to 
the clean surface, therefore oil does not reduce the bite, grip, or 
friction of the air on the water. 
In working with this apparatus it was found very difficult to get 
constant readings. On making a new experiment, after washing 
the glass vessel, some change seemed to take place, and the read- 
ings for the corresponding pressures were. not always the same as 
the previous ones. The general result, however, was that the oil in 
a few cases slightly decreased the readings, but in most cases it 
increased them a little. The difference in the readings in the 
experiments with clean water, seems in some way to be connected 
with the resistance offered by the film where it is attached to the 
sides of the vessel, and the slight increase given in most cases by 
the oily film may possibly be due to the weakening of this film by 
the oil. 
In the experiments described, the surfa’ce film simply circulated 
in the vessel and but little free surface was developed ; if, however, a 
division plate was put across the vessel just deep enough to stop the 
easy circulation of the film, an interesting and curious result was ob- 
tained. When the pressure of the air jet was low, so that it did not 
greatly agitate the surface, and the surface film could circulate 
quickly enough, there was then scarcely any difference between the 
readings given by the clean and by the oily surface. But when 
the pressure rose and the agitation of the water became great, a 
sudden increase in the amount of deflection took place; but this 
increase always occurred at a lower pressure with the oily than 
with the clean film, the explanation of which would seem to be, 
that when greatly agitated by the air new surface film requires to 
be rapidly developed, and, as it gets no aid from the contracting film 
